Pharmacology and therapeutic

implications of current drugs for

type 2 diabetes mellitus
Abd A. Tahrani1,2, Anthony H. Barnett1,2 and Clifford J. Bailey3
Abstract |6[RG|FKCDGVGUOGNNKVWU
6&/KUCINQDCNGRKFGOKEVJCVRQUGUCOCLQTEJCNNGPIGVQ
JGCNVJECTGU[UVGOU+ORTQXKPIOGVCDQNKEEQPVTQNVQCRRTQCEJPQTOCNIN[ECGOKC
YJGTG
RTCEVKECNITGCVN[DGPGHKVUNQPIVGTORTQIPQUGUCPFLWUVKHKGUGCTN[GHHGEVKXGUWUVCKPGFCPF
UCHGV[EQPUEKQWUKPVGTXGPVKQP+ORTQXGOGPVUKPVJGWPFGTUVCPFKPIQHVJGEQORNGZRCVJQIGPGUKU
QH6&/JCXGWPFGTRKPPGFVJGFGXGNQROGPVQHINWEQUGNQYGTKPIVJGTCRKGUYKVJEQORNGOGPVCT[
OGEJCPKUOUQHCEVKQPYJKEJJCXGGZRCPFGFVTGCVOGPVQRVKQPUCPFHCEKNKVCVGFKPFKXKFWCNK\GF
OCPCIGOGPVUVTCVGIKGU1XGTVJGRCUVFGECFGUGXGTCNPGYENCUUGUQHINWEQUGNQYGTKPICIGPVU
JCXGDGGPNKEGPUGFKPENWFKPIINWECIQPNKMGRGRVKFG|TGEGRVQT
).24CIQPKUVUFKRGRVKF[N
RGRVKFCUG|
&22KPJKDKVQTUCPFUQFKWOINWEQUGEQVTCPURQTVGT|
5).6KPJKDKVQTU6JGUG
CIGPVUECPDGWUGFKPFKXKFWCNN[QTKPEQODKPCVKQPYKVJYGNNGUVCDNKUJGFVTGCVOGPVUUWEJCU
DKIWCPKFGUUWNHQP[NWTGCUCPFVJKC\QNKFKPGFKQPGU#NVJQWIJPQXGNCIGPVUJCXGRQVGPVKCN
CFXCPVCIGUKPENWFKPINQYTKUMQHJ[RQIN[ECGOKCCPFJGNRYKVJYGKIJVEQPVTQNNQPIVGTOUCHGV[
JCU[GVVQDGGUVCDNKUJGF+PVJKU4GXKGYYGCUUGUUVJGRJCTOCEQMKPGVKEURJCTOCEQF[PCOKEUCPF
UCHGV[RTQHKNGUKPENWFKPIECTFKQXCUEWNCTUCHGV[QHEWTTGPVN[CXCKNCDNGVJGTCRKGUHQTOCPCIGOGPV
QHJ[RGTIN[ECGOKCKPRCVKGPVUYKVJ6&/YKVJKPVJGEQPVGZVQHFKUGCUGRCVJQIGPGUKUCPFPCVWTCN
JKUVQT[+PCFFKVKQPYGDTKGHN[FGUETKDGVTGCVOGPVCNIQTKVJOUHQTRCVKGPVUYKVJ6&/CPFNGUUQPU
HTQORTGUGPVVJGTCRKGUVQKPHQTOVJGFGXGNQROGPVQHHWVWTGVJGTCRKGU

Type 2 diabetes mellitus (T2DM) is a global epidemic with an
estimated worldwide prevalence of 415 million people in
2015, which is projected to rise to 642 million people
by 2040 (REF. 1). The very considerable health, social and
economic burdens caused by T2DM1–3 present a major
1
challenge to health-care systems around the world.
Centre of Endocrinology,
Diabetes and Metabolism, T2DM is a complex endocrine and metabolic disorder
2nd Floor, Institute of in which the interaction between genetic and environ-
Biomedical Research, mental factors generates a heterogeneous and progressive
University of Birmingham, pathology with varying degrees of insulin resistance and
Birmingham, B15 2TT, UK.
2
Department of Diabetes
dysfunction of pancreatic β cells and α cells, as well as
and Endocrinology, other endocrine disturbances4–14 (FIG. 1). Insulin resist-
Heart of England NHS ance results from deficits in signalling pathways at the
Foundation Trust, level of the insulin receptor and downstream, and T2DM
Birmingham, B9 5SS, UK.
3
emerges when β cells can no longer secrete sufficient
School of Life and Health
Sciences, Aston University, insulin to overcome insulin resistance4,15–17. Overweight
Birmingham, B4 7ET, UK. and obesity are major risk factors for the development of
Correspondence to A.A.T.  insulin resistance4,5,16,18–20.
abd.tahrani@nhs.net Hyperglycaemia is the fundamental biochemical fea-
doi:10.1038/nrendo.2016.86 ture of T2DM, causing oxidative and nitrosative stress
and activation of inflammatory pathways and endothelial
dysfunction, as well as precipitating microvascular com-
plications and contributing to macrovascular disease,
which are major causes of morbidity and mortality 21. The
results of several randomized controlled trials (RCTs)
have demonstrated the short-term and long-term bene-
fits of improving glycaemic control in delaying the onset
and reducing the severity of diabetes-related outcomes,
particularly retinopathy, nephropathy, neuropathy and
cardiovascular disease, and also mortality 22–25. Attaining
normal (or nearly normal) levels of blood glucose (where
practical) is a major aim of T2DM treatment. Several
strategies are available for this purpose: lifestyle changes,
including dietary prudence, weight loss and physical
activity, remain the cornerstones of management, but
because of the progressive nature of T2DM and the dif-
ficulty in maintaining lifestyle changes in the long term,
most patients also require oral therapies and (eventually)
injectable treatments26.
For more than four decades, only two classes of oral
glucose-lowering medications were available (biguanides
and sulfonylureas), but in the past 20 years many more
Key points
• )TGCVGTWPFGTUVCPFKPIQHVJGEQORNGZCPFOWNVKHCEVQTKCNRCVJQIGPGUKUQHV[RG|
FKCDGVGUOGNNKVWU
6&/JCUKPHQTOGFVJGFGXGNQROGPVQHUGXGTCNPGYENCUUGUQH formin and buformin) have been withdrawn because of
INWEQUGNQYGTKPIVJGTCRKGU
• /GVHQTOKPTGOCKPUVJGHKTUVNKPGRJCTOCEQVJGTCR[HQTRCVKGPVUYKVJ6&/YJGTGCU
VJGWUGQHQVJGTYGNNGUVCDNKUJGFCIGPVUUWEJCUUWNHQP[NWTGCUOGINKVKPKFGU
pioglitazone and αINWEQUKFCUGKPJKDKVQTUXCTKGUKPFKHHGTGPVTGIKQPU
• #IGPVUVJCVGPJCPEGKPETGVKPCEVKXKV[
&22KPJKDKVQTUUWRRNGOGPVGPFQIGPQWU Europe in 1957 and in the USA in 1995, and has become
).2
).2TGEGRVQTCIQPKUVUQTKPETGCUGWTKPCT[INWEQUGGNKOKPCVKQP
5).6
KPJKDKVQTUJCXGNQYTKUMQHJ[RQIN[ECGOKCCPFECPCUUKUVYGKIJVEQPVTQN
• 6TGCVOGPVYKVJVYQQTVJTGGCIGPVUYKVJFKHHGTGPVOQFGUQHCEVKQPECPDGTGSWKTGFCU
6&/CFXCPEGUCPFKPUWNKPVJGTCR[KUTGSWKTGFKHQVJGTCIGPVUCTGWPCDNGVQOCKPVCKP
CFGSWCVGIN[ECGOKEEQPVTQN
• )N[ECGOKEVCTIGVUCPFVJGEJQKEGQHINWEQUGNQYGTKPICIGPVUUJQWNFDGEWUVQOK\GF
VQOGGVVJGPGGFUCPFEKTEWOUVCPEGUQHKPFKXKFWCNRCVKGPVUYJKEJEQWNFDGHCEKNKVCVGF
D[HWVWTGFGXGNQROGPVUKPRJCTOCEQIGPQOKEU
• #NVJQWIJVJGDCNCPEGQHDGPGHKVUCPFTKUMUHQTFKHHGTGPVCIGPVUXCTKGUDGVYGGP
KPFKXKFWCNRCVKGPVUGCTN[GHHGEVKXGCPFUWUVCKPGFIN[ECGOKEEQPVTQNFGNC[UVJGQPUGV
CPFTGFWEGUVJGUGXGTKV[QHJ[RGTIN[ECGOKCTGNCVGFEQORNKECVKQPU
treatment options have been introduced26,27 (TABLE 1).
In this Review, we provide an evaluation of the thera-
pies available for the management of hyperglycaemia in
patients with T2DM.
Glycaemic control and targets in T2DM
The treatment needs of patients with T2DM, and the
responses to treatments, are highly variable, reflecting the
complexity and variability of the pathogenic process28,29,
so decisions must be made for each patient regarding the
choice of therapy and glycaemic targets. Factors for con-
sideration include patient age, weight, duration of T2DM,
risk of hypoglycaemia, cardiovascular risk, concomitant
treatments, presence of complications and concomitant
life-limiting illness. Other aspects, which are more diffi-
cult to quantify in clinical practice, include the reserve
capacity for insulin secretion, genetic factors that might
affect responses to therapies, the risk of developing future
complications and the rate of disease progression30.
The long-term benefits of intensive glycaemic control
on T2DM-related complications and mortality are well
known, particularly when initiated promptly after diag-
nosis in young patients who do not yet have comorbid
complications22–25. However, intensive glycaemic control
is not without risks, such as hypoglycaemia, weight gain
and possible cardiovascular events and mortality in high-
risk individuals. These risks might relate, at least in part,
to the choice of glycaemic target and medications22,31–36,
so an individualized management strategy is prefera-
ble36. The difficulty lies in the identification of patients in
whom the risks associated with intensive glycaemic con-
trol outweigh the benefits. Stringent glycaemic control is
not advised in elderly patients or in those with advanced
disease, long T2DM duration or established cardiovascu-
lar disease27,36. An HbA1c target of 7% is commonly given
in guidelines, but a lower target might be appropriate for
newly diagnosed, young patients with T2DM and no
complications, and a higher target might be more real-
istic for an elderly or frail patient with a long duration of
disease and established complications.
Biguanides
The only biguanide available in clinical practice is met-
formin (dimethylbiguanide)37. Other biguanides (phen-
risks of lactic acidosis38. Biguanides were derived from
the guanidine-rich herb Galega officinalis (French lilac),
which was used in traditional medicine in Europe37,39.
Metformin was introduced into clinical practice in
the most prescribed agent for T2DM worldwide37,39.
Mechanism of action
Metformin enters cells mainly via solute carrier
family 22 member 1 (also known as organic cation
transporter 1 (hOCT1)) and exerts multiple insulin-
dependent and insulin-independent actions according
to the level of drug exposure and the control of nutri-
ent metabolism within different tissues28,37,40–42 (FIG. 2).
During treatment, the gut is exposed to high concen-
trations of metformin 42, which interrupt the mito-
chondrial respiratory chain at complex I, and increase
glucose utilization, anaerobic glycolysis and lactate
production; some of the lactate can be converted back
to glucose in the liver 43. Lactate–glucose turn over
causes energy dissipation, which might contribute to
the weight neutrality (lack of weight gain or weight
loss) observed in metformin-treated patients28,42. In the
liver, metformin increases insulin signalling, reduces
glucagon action and reduces gluconeogenesis and
glycogenolysis28. Metformin can inhibit the mitochon-
drial redox shuttle enzyme glycerol-3-phosphate dehy-
drogenase, altering the hepatocellular redox state and
resulting in reductions in the ATP:AMP ratio, hepatic
gluconeogenesis and the conversion of lactate and
glycerol to glucose, and activation of AMP-activated
protein kinase (AMPK)44. In addition, metformin treat-
ment results in a shift toward the utilization of glucose
relative to fatty acids as a cellular source of energy in
the liver 37. In muscle, metformin promotes insulin-
mediated glucose uptake via solute carrier family 2,
facilitated glucose transporter member 4 (GLUT-4)28.
As delayed-release formulations of metformin have
achieved similar efficacies at lower doses compared with
‘regular’ formulations, it seems that the gut is a major site
of metformin action at therapeutic doses45. Metformin
can increase circulating levels of glucagon-like peptide-1
(GLP-1) from pretreatment levels, even in the absence
of an oral glucose load and in individuals with and
without T2DM46–50, by mechanisms that could include
inhibition of sodium-dependent bile-acid transporters,
which increase the availability of ileal bile acids to acti-
vate G-protein coupled bile acid receptor 1 (commonly
known as TGR5) on enteroendocrine L cells. Compared
with placebo, metformin reduces the activity of dipepti-
dyl peptidase 4 (DPP-4)46. Relative to pretreatment lev-
els, metformin increases GLP-1 secretion in response
to an oral glucose load, via muscarinic (M3) and gas-
trin-releasing peptide receptor (GRP-R)-dependent
pathways47–51. In mice, metformin stimulates expres-
sion of GLP-1 receptor (Glp-1r) on pancreatic β cells,
mediated by peroxisome proliferator-activated receptor
Factors that contribute to defective
blood glucose control in T2DM
Muscle Pancreas Brain Liver
Impaired glucose • Impaired insulin secretion Neurotransmitter Excess glucose
uptake, storage and and loss of β-cell mass dysfunction and production and
metabolism • Inappropriately raised abnormalities of excess lipid storage
glucagon secretion circadian rhythm

Gut Adipose tissue Kidney

Disturbances of Increased lipid storage, Increased glucose
incretin function defective adipokine reabsorption
and the microbiome RTQFWEVKQPKPȯCOOCVKQP

Insulin GLP-1 receptor agonists • Pramlintide Metformin

• Bromocriptine
• Sulfonylureas
• Meglitinides

• Metformin Thiazolidinediones SGLT2 inhibitors

• DPP-4 inhibitors
• Colesevelam
• α-glucosidase inhibitors

Lifestyle, diet
Blood glucose
and exercise

6KUUWGURGEKȮEGȭGEVUKP6&/
Weight gain Weight neutral Weight loss

(KIWTG| Sites of action of glucose-lowering agents. /WNVKRNGIGPGVKECPFGPXKTQPOGPVCNHCEVQTUIKXGTKUGVQV[RG|

FKCDGVGUOGNNKVWU
6&/VJTQWIJKPUWNKPTGUKUVCPEGYKVJRCPETGCVKEβEGNNHCKNWTG1XGTYGKIJVCPFQDGUKV[EQPVTKDWVGVQ
KPUWNKPTGUKUVCPEGKPCUUQEKCVKQPYKVJKPETGCUGFKPHNCOOCVQT[UKIPCNUCPFFKUVWTDGFNKRKFJQOGQUVCUKUQHVGPRTGEGFKPIVJG
QPUGVQHJ[RGTIN[ECGOKCD[OCP[[GCTUCPFGPJCPEKPIECTFKQXCUEWNCTTKUM9JGPKPUWNKPUGETGVKQPKUPQNQPIGTUWHHKEKGPV
VQQXGTEQOGKPUWNKPTGUKUVCPEGINWEQUGKPVQNGTCPEGRTQITGUUGUVQ6&/WUWCNN[CEEQORCPKGFD[RCPETGCVKEαEGNN
F[UHWPEVKQPVJCVGNGXCVGUINWECIQPUGETGVKQPTGFWEGFRTCPFKCNUGETGVKQPQTCEVKXKV[QHKPETGVKPJQTOQPGUUWEJCU
INWECIQPNKMGRGRVKFG
).2CNVGTCVKQPUVQVJGIWVOKETQDKQOGCPFFKUVWTDCPEGUQHPGWTCNCEVKXKVKGUEQPVTQNNKPI
JWPIGTsUCVKGV[CPFVJGEKTECFKCPTGIWNCVKQPQHINWEQUGJQOGQUVCUKU+PUWNKPUWNHQP[NWTGCUCPFOGINKVKPKFGUCTGCUUQEKCVGF
YKVJTKUMQHJ[RQIN[ECGOKC&22FKRGRVKF[NRGRVKFCUG|5).6UQFKWOINWEQUGEQVTCPURQTVGT|

(PPAR) α49. The effect of metformin on GLP-1 might Pharmacodynamics

contribute to its weight-neutral effect and to reduction
in hepatic glucose output by inhibiting glucagon secre-
tion46–48. Metformin also affects the circadian control of
glucose metabolism in liver and muscle42. Metformin-
induced AMPK activation results in phosphorylation
of casein kinase I, which leads to degradation of the
circadian clock component mPer2, thereby increasing
expression of the CLOCK and BMAL1 circadian genes
and causing phase advance in the circadian rhythm in
treated rodents, compared with untreated controls52,53.
The results of a study involving mice showed that met-
formin causes phase advance in the liver, but phase delay
in muscle53, and the effects of metformin on circadian
rhythm are blocked in mice with knock-out of Prkaa2,
the gene encoding AMPK subunit α2 (REF. 52).
Pharmacokinetics
Metformin has an oral bioavailability of 40–60% and a
plasma half-life of 4–9 h, and is eliminated unchanged
in the urine mostly via tubular secretion rather than
glomerular filtration28,54.
Metformin is widely used as a first-line pharmacotherapy
in patients with T2DM, because of its efficacy, long-term
safety record, low risk of hypoglycaemia, weight neutral-
ity and favourable effect on vascular disease36. Metformin
treatment typically leads to a reduction in fasting plasma
glucose (FPG) by 2–4 mmol/l and HbA1c by 1–2%, largely
independent of age, weight and T2DM duration as long as
some residual β-cell function remains28,39. In the 10-year
follow-up data from the UK Prospective Diabetes Study
(UKPDS), patients who received metformin had signifi-
cant risk reductions for any diabetes-related end point of
21% (P = 0.01), diabetes-related death of 30% (P = 0.01)
and myocardial infarction of 33% (P = 0.005) compared
with overweight patients in the conventional therapy
group23,28,55. Metformin might also be associated with a
reduction in the risk of cancer in patients with T2DM,
particularly prostate, pancreas and breast cancer 28,42.
The progressive nature of T2DM can require the addi-
tion of other glucose-lowering treatments (including insu-
lin) to metformin15,36,56. Many fixed-dose combinations of
drugs that include metformin are, therefore, available.
6CDNG| 5WOOCT[QHEWTTGPVN[CXCKNCDNGINWEQUGNQYGTKPIVTGCVOGPVUKPRCVKGPVUYKVJV[RG|FKCDGVGUOGNNKVWU,36
Class and Dosing Mechanism of Physiological Glucose- Advantages Disadvantages Cardiovascular Cost||
examples action effects lowering safety
efficacy
Sulfonylureas (1956)*
• )NKENC\KFG‡ • 1& $KPFVQ574QP +PETGCUG *KIJ )QQFNQPIVGTO • *[RQIN[ECGOKC %QPHNKEVKPI .QY
• )NKRK\KFG • $& β|EGNNUTGUWNVKPI KPUWNKP UCHGV[ • 9GKIJVICKP TGUWNVUHTQO
• )NKOGRKTKFG KPENQUWTGQH UGETGVKQP • 0GGFHQT5/$) FCVCDCUG
• )N[DWTKFG K#62EJCPPGNU • 0GGFHQTFQUG UVWFKGUPQ

INKDGP FGRQNCTK\CVKQP VKVTCVKQP CFXGTUG
ENCOKFG CPFECNEKWO QWVEQOGUKP
KPHNWZ KPVGTXGPVKQPCN
UVWFKGU
Biguanides (1957)*
• /GVHQTOKP • 1& • #EVKXCVG • 4GFWEG *KIJ • )QQF • )CUVTQKPVGUVKPCN 4GFWEG .QY
• /GVHQTOKP • $& #/2- JGRCVKE NQPIVGTO CFXGTUGGHHGEVU ECTFKQXCUEWNCT
UNQYTGNGCUG • +ORTQXG INWEQUG UCHGV[ • /WNVKRNGRQUUKDNG FKUGCUG
EGNNWNCTKPUWNKP QWVRWV • 9GKIJVPGWVTCN EQPVTCKPFKECVKQPU
UKIPCNNKPI • +ORTQXG • .QYTKUMQH GURGEKCNN[TGPCN
• 4GFWEG KPUWNKP J[RQIN[ECGOKC KORCKTOGPVCPF
TGURKTCVQT[ UGPUKVKXKV[ J[RQZCGOKC
EJCKPCEVKXKV[ • +PETGCUG
• #NVGTIWV ).2NGXGNU
INWEQUGs
NCEVCVG
OGVCDQNKUO

α-Glucosidase inhibitors (1995)*

• #ECTDQUG 7RVQ +PJKDKV 5NQYKPVGUVKPCN /QFGUV 9GKIJVPGWVTCN )CUVTQKPVGUVKPCN 7PMPQYP /QFGTCVG
• /KINKVQN 6&5 αINWEQUKFCUGKP ECTDQJ[FTCVG CFXGTUGGHHGEVU RTGNKOKPCT[
• 8QINKDQUG YKVJ VJGIWV FKIGUVKQP GXKFGPEGQH
OGCNU CPFFGNC[ DGPGHKVU
CDUQTRVKQP

Meglitinides (1997)*
• 0CVGINKPKFG 9KVJ • $KPFVQ574 +PETGCUG +PVGT • 4CRKFQPUGV • 9GKIJVICKP %CTFKQXCUEWNCT /QFGTCVG
• 4GRCINKPKFG OGCNU QPβ|EGNNU KPUWNKP OGFKCVG UJQTVFWTCVKQP • *[RQIN[ECGOKC FKUGCUGPQV
• #EVKQPU UGETGVKQP VQJKIJ • 5WKVCDNGHQT • 0GGFHQT5/$) CFXGTUGN[
OQTGTCRKF RTCPFKCNWUG
NGUUVJCPYKVJ CHHGEVGF
CPFUJQTVGT UWNHQP[NWTGCU
FWTCVKQPVJCP
UWNHQP[NWTGCU
Thiazolidinediones (1997)*
• 2KQINKVC\QPG 1& 22#4γCIQPKUVU • +PETGCUG *KIJ • .QYTKUMQH • 7PTGUQNXGF • 1GFGOCCPF .QY
• 4QUKINKVC\QPG§ KPUWNKP J[RQIN[ECGOKC NQPIVGTOUCHGV[ KPETGCUGFTKUM
UGPUKVKXKV[ • /KIJVTGFWEG • (TCEVWTGU QHJGCTVHCKNWTG
• 4GFWEGHTGG DNQQFRTGUUWTG • 9GKIJVICKP • &GDCVGF
HCVV[CEKF • 2QUUKDNG • 1GFGOCCPFJGCTV GHHGEVQP
TGNGCUG GHHGEVQP HCKNWTG ECTFKQXCUEWNCT
PQPCNEQJQNKE FKUGCUG
UVGCVQJGRCVKVKU • 2KQINKVC\QPG
TGFWEGF
EQORQUKVG
GPFRQKPV

DPP-4 inhibitors (2006)*

• 5KVCINKRVKP • 1& +PJKDKV&22 )NWEQUG +PVGT • 9GKIJVPGWVTCN • 7PMPQYP 0QKPETGCUGKP *KIJ
• 8KNFCINKRVKP‡ • $& CEVKXKV[KPETGCUG FGRGPFGPV OGFKCVG • .QYTKUMQH NQPIVGTOUCHGV[ ECTFKQXCUEWNCT
• 5CZCINKRVKP GPFQIGPQWU KPETGCUG J[RQIN[ECGOKC • +PETGCUGFTKUMQH FKUGCUGTKUM
• .KPCINKRVKP KPETGVKPNGXGNU KPKPUWNKP
WPNGUU RCPETGCVKVKU TGRQTVGFGZEGRV
• #NQINKRVKP UGETGVKQPCPF EQODKPGFYKVJ • 2QUUKDNGKPETGCUGF KPETGCUGF
KPJKDKVKQP UWNHQP[NWTGC TKUMQHNKXGT JQURKVCNK\CVKQP
QHINWECIQP • 2QUUKDNG F[UHWPEVKQPYKVJ YKVJJGCTV
UGETGVKQP DGPGHKVQP XKNFCINKRVKP HCKNWTGYKVJ
βEGNNUWTXKXCN UCZCINKRVKP
6CDNG
EQPV| 5WOOCT[QHEWTTGPVN[CXCKNCDNGINWEQUGNQYGTKPIVTGCVOGPVUKPRCVKGPVUYKVJV[RG|FKCDGVGUOGNNKVWU,36
Class and Dosing Mechanism of Physiological Glucose- Advantages Disadvantages Cardiovascular Cost||
examples action effects lowering safety
efficacy
SGLT2 inhibitors (2012)*
• %CPCINKHNQ\KP 1& +PJKDKV5).6 +PETGCUG +PVGT • 9GKIJVNQUU • 7PMPQYP 'ORCINKHNQ\KP *KIJ
• &CRCINKHNQ\KP VTCPURQTVGTUKP WTKPCT[ OGFKCVGVQ • $NQQFRTGUUWTG NQPIVGTOUCHGV[ TGFWEGU
• 'ORCINKHNQ\KP RTQZKOCNTGPCN INWEQUG JKIJ TGFWEVKQP • #UUQEKCVKQP ECTFKQXCUEWNCT
VWDWNGU GZETGVKQP • .QYTKUMQH YKVJIGPKVCNCPF FKUGCUG
J[RQIN[ECGOKC RQUUKDN[WTKPCT[

WPNGUU VTCEVKPHGEVKQPU
EQODKPGF • 1UOQVKEFKWTGUKU
YKVJKPUWNKPQT RQUUKDNGTKUMQH
UWNHQP[NWTGC J[RQVGPUKQPCPF
• 2QUUKDNG HCNNU
UWUVCKPGF • 2QUUKDNGKPETGCUGF
HbAE TKUMQHHTCEVWTGU
TGFWEVKQP • 5OCNNKPETGCUGF
TKUMQHFKCDGVKE
MGVQCEKFQUKU

Dopamine-2 agonist (2009)*

$TQOQETKRVKPG 1& #EVKXCVGU • 5WRRTGUUKQP /QFGUV • 9GKIJVPGWVTCN • &K\\KPGUU 4GFWEGU *KIJ
SWKEMTGNGCUG J[RQVJCNCOKE QHJGRCVKE • .QYTKUMQH • 0CWUGC ECTFKQXCUEWNCT
FQRCOKPG INWEQUG J[RQIN[ECGOKC • (CVKIWG FKUGCUGTKUM
TGEGRVQTU QWVRWV
• +PETGCUGU
INWEQUG
FKURQUCN

Bile-acid sequestrant (2008)*

%QNGUGXGNCO • 1& • +PETGCUGU 2QUUKDN[ /QFGUV • .QYTKUMQH • %QPUVKRCVKQP 4GFWEGUTKUMQH *KIJ
• $& JGRCVKE TGFWEGU J[RQIN[ECGOKC • +PETGCUGU ECTFKQXCUEWNCT
DKNGUCNV JGRCVKE • 9GKIJVPGWVTCN VTKIN[EGTKFGU FKUGCUG
RTQFWEVKQP INWEQUGQWVRWV • 4GFWEGU.&. • %QWNFCHHGEV
NKEGPUGFCUEJQ
• +PETGCUGU CPFKPETGCUGU EJQNGUVGTQN CDUQTRVKQPQHUQOG NGUVGTQNNQYGT
).2 KPETGVKP KPETGCUGU*&. FTWIU KPIVTGCVOGPV
UGETGVKQP UGETGVKQP EJQNGUVGTQN
• #EVKXCVGUNKXGT
HCTPGUQKF
TGEGRVQTU

Insulin (1920s)*
• 4CRKFCEVKPI 1&VQ &KTGEVN[CEVKXCVG • +PETGCUG *KIJ • +PLGEVCDNG • 9GKIJVICKP 1PIQKPI 8CTKCDNG

CURCTVNKURTQ 3&5 VJGKPUWNKP INWEQUG • 5WUVCKPGF • *[RQIN[ECGOKC FGDCVG
INWNKUKPG TGEGRVQT FKURQUCN IN[ECGOKE • 0GGFHQT5/$) KPETGCUGFTKUM
• 5JQTVCEVKPI • 4GFWEG KORTQXGOGPVU • (NWKFTGVGPVKQP PQVUJQYPKP

JWOWNKP5 JGRCVKE EQORCTGFYKVJ 4%6U
KPUWOCPTCRKF INWEQUG QVJGTCIGPVU
CEVTCRKF QWVRWV
• +PVGTOGFKCVG • &GETGCUG
CEVKPI NKRQN[UKU

KPUWNCVCTF
JWOWNKP+
KPUWOCPDCUCN
• .QPICEVKPI

INCTIKPG
FGVGOKT
FGINWFGE
• $KRJCUKE
RTGOKZGF
6CDNG
EQPV| 5WOOCT[QHEWTTGPVN[CXCKNCDNGINWEQUGNQYGTKPIVTGCVOGPVUKPRCVKGPVUYKVJV[RG|FKCDGVGUOGNNKVWU,36
Class and Dosing Mechanism of Physiological Glucose- Advantages Disadvantages Cardiovascular Cost||
examples action effects lowering safety
efficacy
GLP-1RAs (2005)*
• 'ZGPCVKFG • 1& #EVKXCVGVJG • )NWEQUG *KIJ • 9GKIJVNQUU • +PLGEVCDNG • 2QUUKDNG *KIJ
• .KTCINWVKFG • $& ).2TGEGRVQT FGRGPFGPV • .QYTKUMQH • )CUVTQKPVGUVKPCN DGPGHKEKCN
• .KZKUGPCVKFG • 39 KPETGCUG J[RQIN[ECGOKC CFXGTUGGHHGEVU GHHGEVKPPQP
• #NDKINWVKFG KPKPUWNKP
WPNGUU • 7PMPQYP TCPFQOK\GF
• &WNCINWVKFG UGETGVKQP EQODKPGFYKVJ NQPIVGTOUCHGV[ UVWFKGU
CPF UWNHQP[NWTGC • 7PEQPHKTOGF • .KZKUGPCVKFG
KPJKDKVKQP • 2QUUKDNGGHHGEV KPETGCUGFTKUMQH FKFPQVCNVGT
QHINWECIQP QPβEGNN RCPETGCVKVKU ECTFKQXCUEWNCT
UGETGVKQP UWTXKXCN FKUGCUGTKUMKP
• 4GFWEG • 2QUUKDNG 4%6
RQUVRTCPFKCN UWUVCKPGF
INWEQUG HbAE
GZETGVKQP TGFWEVKQP
• +PETGCUG
UCVKGV[
Amylin analogue (2005)*
2TCONKPVKFG‡ 6&5 5[PVJGVKE • 4GFWEGU /QFGUV • 9GKIJVNQUU • +PLGEVCDNG 7PMPQYP *KIJ
UQNWDNG INWECIQP • 4GFWEGF • 7PMPQYP
CPCNQIWGQH UGETGVKQP KPUWNKPFQUG NQPIVGTOUCHGV[
JWOCPCO[NKP • +PETGCUGU • +PETGCUGFTKUMQH
UCVKGV[ J[RQIN[ECGOKC
• 5NQYUICUVTKE • 1PN[WUGFYKVJ
GORV[KPI OGCNVKOGKPUWNKP
;GCTQHKPVTQFWEVKQPQHHKTUVKPENCUU‡0QVCXCKNCDNGKPCNNTGIKQPU§&KUEQPVKPWGFKP'WTQRG||%QUVKUDCUGFQPNQYGUVRTKEGFOGODGTQHVJGENCUU#/2-ʹ
#/2CEVKXCVGFRTQVGKPMKPCUG$&VYKEGFCKN[&22FKRGRVKF[NRGRVKFCUG|).2INWECIQPNKMGRGRVKFG|).24#).2TGEGRVQTCIQPKUV1&QPEGFCKN[
22#4γRGTQZKUQOGRTQNKHGTCVQTCEVKXCVGFTGEGRVQT|γ3&5HQWTVKOGUFCKN[39QPEGYGGMN[4%6TCPFQOK\GFEQPVTQNNGFVTKCN5).6UQFKWOINWEQUG
Safety and adverse effects
The main adverse effects of metformin treatment are
abdominal discomfort and other gastrointestinal effects,
including diarrhoea37. Symptoms can diminish if the dose
is reduced, but around 10% of patients cannot tolerate the
drug at any dose37, possibly because of variants of hOCT1
that lead to an increased concentration of metformin
in the intestine57. The risk of metformin intolerance
(defined as patients who stop metformin within the first
6 months of treatment) is increased by concomitant use
of drugs that inhibit hOCT1 activity (including tricyclic
antidepressants, citalopram, proton-pump inhibitors,
verapamil, diltiazem, doxazosin, spironolactone, clopi-
dogrel, rosiglitazone, quinine, tramadol and codeine;
OR 1.63, 95% CI 1.22–2.17, P = 0.001) or the presence of
two alleles of SLC22A1 associated with reduced function
of hOCT1 rather than one allele or no deficient allele
(OR 2.41, 95% CI 1.48–3.93, P <0.001)57.
Metformin is contraindicated in patients with advanced
chronic kidney disease (CKD), notable liver disease or con-
ditions that might predispose to hypoxia or reduced tissue
perfusion. However, observational and database studies
indicate that advantage can be taken of the broad thera-
peutic index with metformin38,58,59, and careful attention to
dose has enabled its use even in patients with cardiovas-
cular disease (including mild-to-moderate heart failure38,60
and chronic obstructive pulmonary disease61). Adjusting
the dose and monitoring renal function to ensure adequate
elimination are important considerations, and metformin
therapy should be stopped if hypoxaemia occurs62,63.
Results of the UKPDS showed that, compared with
sulfonylureas and insulin in patients with obesity and
newly diagnosed T2DM, metformin use was associated
with significantly reduced rates of myocardial infarction,
stroke and all-cause mortality (by 39%, 41% and 36%,
respectively)64,65. The 10-year follow-up of the UKPDS
showed that the reductions in myocardial infarction
and mortality persist 23. Database analyses have consist-
ently provided corroborating evidence for this effect 65.
Increasing levels of use of statins and renal-protective
medications make it difficult to assess the effect of met-
formin on cardiovascular disease65, although several RCTs
are ongoing to assess this effect65.
Sulfonylureas
Sulfonylureas were developed as variants of sulfonamides
after the latter were reported to cause hypoglycaemia37,66.
Sulfonylureas are classified as first-generation (such as tol-
butamide and chlorpropamide) and second-generation
(such as glibenclamide (glyburide), gliclazide, glipizide
and glimepiride)37; the second-generation drugs have
greater potency, enabling treatment with lower doses.
Mechanism of action
Sulfonylureas act directly on pancreatic β cells by binding
to the cytosolic face of ATP-binding cassette sub-family C
member 8 (also known as sulfonylurea receptor 1 (SUR1)),
which is part of the Kir6.2 ATP-sensitive potassium chan-
nel37,67. Binding closes the Kir6.2 channel, preventing
potassium efflux and depolarizing the plasma membrane.
 High metformin Liver Lower metformin
exposure in the Glucagon exposure in liver
intestine Insulin and muscle
Glucagon
Metformin receptor Glucose
↓ PTP1B Muscle
GLUT2
↓ Other phosphatases?
GLUT4
hOCT1 Insulin receptor
ATP
cAMP
PIP2 Translocation
↓ GPD-M
IRS1/2 PI3K ↓ G6Pase
↓ Complex 1
PIP3 ↓ FBPase
Mitochondrion
PKB PDK1/2 Pyruvate
↓ ATP:AMP
LKB ↓ Gluconeogenesis
AMPK

Glucose
Small vesicle uptake
mTORC1

Anticancer
GȭGEVU! ACC MAPK cascade

Protein synthesis, ↓ Gluconeogenesis Transcription of genes ↑ Glucose uptake,

growth and proliferation ↓ Lipid synthesis involved in energy metabolism
↑ Lactate production storage and expenditure and glycogenesis

(KIWTG| Intracellular actions of metformin. /GVHQTOKPCNVGTUPWVTKGPVOGVCDQNKUOVJTQWIJKPUWNKPFGRGPFGPVCPF

KPUWNKPKPFGRGPFGPVGHHGEVUVJCVXCT[YKVJVJGCOQWPVQHFTWIGZRQUWTGCPFVJGCEVKXKV[QHKPUWNKPYKVJKPFKHHGTGPVVKUUWGU
6JGKPVGUVKPGKUGZRQUGFVQJKIJNGXGNUQHOGVHQTOKPYJKEJJCXGKPUWNKPKPFGRGPFGPVGHHGEVUYJGTGCUNKXGTCPFOWUENGCTG
GZRQUGFVQNQYGTEQPEGPVTCVKQPUQHOGVHQTOKPVJCVKPHNWGPEGVJGOGVCDQNKEGHHGEVUQHKPUWNKP/GVHQTOKPECPKORTQXG
KPUWNKPUGPUKVKXKV[XKCGHHGEVUQPKPUWNKPTGEGRVQTUKIPCNNKPICPFRQUVTGEGRVQTUKIPCNNKPIRCVJYC[UQHKPUWNKPCEVKQP
/GVHQTOKPECPCNVGTEGNNWNCTPWVTKGPVOGVCDQNKUOCPFGPGTI[RTQFWEVKQPKPFGRGPFGPVN[QHKPUWNKPXKCUWRRTGUUKQPQHVJG
OKVQEJQPFTKCNTGURKTCVQT[EJCKPCPFCEVKXCVKQPQHʹ#/2CEVKXCVGFRTQVGKPMKPCUG
#/2-#%%CEGV[N%Q#ECTDQZ[NCUG
($2CUGHTWEVQUGDKURJQURJCVCUG)2CUGINWEQUGRJQURJCVCUG).76INWEQUGVTCPURQTVGTKUQHQTOJ1%6
QTICPKEECVKQPVTCPURQTVGT|+45KPUWNKPTGEGRVQTUWDUVTCVG.-$UGTKPGVJTGQPKPGsRTQVGKPMKPCUG56-/#2-
OKVQIGPCEVKXCVGFRTQVGKPMKPCUGO614OCOOCNKCPVCTIGVQHTCRCO[EKP2&-RJQURJQKPQUKVKFGFGRGPFGPVRTQVGKP
MKPCUG2+-RJQURJCVKF[NKPQUKVQNMKPCUG2+2RJQURJCVKF[NKPQUKVQNDKURJQURJCVG2+2RJQURJCVKF[NKPQUKVQN
VTKURJQURJCVG2-$RTQVGKPMKPCUG|$)2&/IN[EGTQNRJQURJCVGFGJ[FTQIGPCUGOKVQEJQPFTKCN

This opens local voltage-dependent calcium channels,
increasing the influx of calcium and activating calci-
um-dependent signalling proteins, leading to insulin
exocytosis (FIG. 3). In vitro studies show that persistent
exposure to sulfonylureas for several days can desensi-
tize β cells and reduce the insulin-secretory response.
However, studies in patients with T2DM have shown
that a 25% increase in 24-h insulin secretion with the
sulfonylurea glibenclamide is maintained for 6–10 weeks,
although efficacy usually declines after 6–12 months of
sulfonylurea therapy during clinical trials68.
Pharmacokinetics
Sulfonylureas vary considerably in their pharmacoki-
netic properties37,68–70 (see Supplementary informa-
tion S1 (table)). They have high bioavailability and reach
peak plasma concentrations within 1.5–4.0 h68. They
are metabolized in the liver to varying extents to form a
number of active and inactive metabolites that are elimi-
nated along with unchanged drug via the bile and urine;
caution is needed when treating patients with hepatic
and/or renal impairment37. Half-lives are <10 h for some
sulfonylureas, but extend to >24 h for others. Therapeutic
effects are exerted for much longer than is indicated by
the half-life if active metabolites are formed (as they are
with glimepiride, glibenclamide and chlorpropamide)68.
In general, first-generation sulfonylureas should be
avoided in patients with CKD stages 3 or 4 or those who
are undergoing dialysis, in whom gliclazide and glip-
izide are suitable without extensive dose adjustment 71–73.
Glimepiride is an option for patients with CKD but not
receiving dialysis, on the proviso of low-dose initiation
and careful titration71,73.
More than 90% of sulfonylureas in the circulation
are bound to plasma proteins, which can lead to inter-
actions with other protein-bound drugs such as salicy-
lates, sulfonamides and warfarin37,68. Some medications
potentiate the glucose-lowering effects of sulfonylureas
by inhibition of their hepatic metabolism (for exam-
ple, some antifungals and monoamine oxidase inhibi-
tors), displacing them from binding to plasma proteins
(for example, coumarins, NSAIDs and sulfonamides),
inhibiting their excretion (for example, probenecid) or
antagonizing their mechanism of action (for example,
diazoxide and other KATP-channel openers)37. Drugs
such as rifampicin that induce sulfonylurea metabolism
inhibit glucose-lowering by sulfonylureas37.
Altered sulfonylurea formulations can enable rapid
onset of action (as is the case with micronized gliben-
clamide) or prolonged activity (for example, ‘Glipizide
Extended Release’ and ‘Gliclazide Modified Release’)
while maintaining glucose-lowering efficacy 37,74–76.
Pharmacodynamics
As monotherapy, sulfonylureas can lead to reductions in
FPG by 2–4 mmol/l and HbA1c by 1–2%28,37,68,70. However,
the failure rates of sulfonylureas as monotherapy are
greater than those of metformin or rosiglitazone15.
Sulfonylureas can be used as first-line treatment options
in patients who are intolerant of metformin, and can be
used in combination with most other glucose-lowering
medications, except meglitinides, which have a similar
mechanism of action28,37. The size and durability of the
response to sulfonylureas is positively associated with
the reserve of β-cell function37.
Safety and adverse effects
Hypoglycaemia and weight gain are the main adverse
effects associated with sulfonylureas. Weight gain of
1–4 kg that stabilizes after about 6 months is common
following drug initiation28. Weight gain is probably
related to the anabolic effect of the increased insulin
levels and reduction of glycosuria27,28,56.
Hypoglycaemia has been reported in 20–40% of
patients receiving sulfonylureas, and severe hypoglycae-
mia (requiring third-party assistance) occurs in 1–7%
of patients28,37,77, depending on the population, the defi-
nition of hypoglycaemia and the type and pharmaco-
kinetics of the sulfonylurea74. In a study involving six UK
secondary care centres, self-reported hypoglycaemia
prevalence was 39% (95% CI 30–49%), similar to that in
patients with T2DM treated with insulin for <2 years77.
The prevalence of self-reported severe hypoglycaemia
was 7% (95% CI 3–13%)77. Continuous glucose moni-
toring (CGM) showed that 22% (95% CI 15–31%) of
patients had at least one episode of interstitial glucose
<2.2 mmol/l, similar to patients with T2DM treated
with insulin for <2 years77. These results confirmed that
the use of long-acting sulfonylureas with active metab-
olites is especially associated with hypoglycaemia28,37,
and that the elderly, those living alone and those with
renal or liver impairment, as well as car drivers, require
extreme caution during treatment with sulfonylureas, as
do those prescribing these drugs28,37. Education and glu-
cose self-monitoring are essential in patients receiving
sulfonylureas; the results of an RCT78 involving patients
receiving Gliclazide Modified Release showed that
self-monitoring of blood glucose reduced both the risk of
symptomatic hypoglycaemia and the reduction in HbA1c,
compared with no monitoring.
The cardiovascular safety of sulfonylureas is
controversial. In the 1970s, the University Group
Diabetes Program raised concerns regarding increased
cardiovascular disease risk with tolbutamide79, and since
then many database studies, mostly retrospective, have
suggested that sulfonylureas (particularly glibenclamide)
are associated with less benefit than metformin against
cardiovascular disease in patients with T2DM65. However,
the results of RCTs such as UKPDS, ADVANCE and
ACCORD did not show an increase in cardiovascular
mortality or morbidity in sulfonylurea-treated patients65.
The ongoing CAROLINA study 80 comparing linaglitpin
with glimepiride in patients with T2DM might help to
define the cardiovascular safety of these drugs.
Meglitinides
The two main meglitinides (or glinides) are nateglin-
ide and repaglinide. The class takes its name from the
meglitinide moiety of glibenclamide, which exerts an
insulin-releasing effect independently of the sulfonyl
moiety 26,28,81.
Mechanism of action
Meglitinides bind to the benzamido site of SUR1 on
β cells. This site is separate from the sulfonyl-binding
site, but meglitinide binding has a similar effect to sulfo-
nylurea binding on the Kir6.2 channels37 (FIG. 3). However,
the relatively rapid onset and short duration of action of
meglitinides suits their use as prandial glucose-lowering
agents37.
Pharmacokinetics
Repaglinide is almost completely absorbed, with peak
plasma concentrations after about 1 h. Repaglinide
binds to proteins in the circulation, and is rapidly
metabolized in the liver (mostly by cytochrome P450
3A4 (CYP3A4)), producing inactive metabolites that are
mostly excreted in the bile. A plasma half-life of around
1 h37,82,83 makes it suitable for patients with poor renal
function. Taken approximately 15 min before a meal,
repaglinide produces a prompt insulin response that
lasts 4–6 h37. Bioavailability is unaffected by the inges-
tion of food. Repaglinide concentrations are positively
affected by co-treatment with drugs that inhibit CYP3A4
(such as ketoconazole, antibacterial agents, steroids and
cyclosporine), and negatively affected by drugs that
induce CYP3A4 (such as rifampicin, carbamazepine
and barbiturates)83,84.
Nateglinide has a slightly faster onset and shorter
duration of action (3–5 h) than repaglinide, but is also
protein-bound in the circulation and metabolized in the
liver by CYP3A4, producing metabolites that are mostly
excreted in the urine37,83.
Pharmacodynamics
Repaglinide (0.5–4.0 mg) or nateglinide (60–180 mg)
taken before meals produces dose-dependent increases
in insulin concentrations and reduces postprandial and
fasting hyperglycaemia37. Meglitinides are well-suited
to patients with irregular meal patterns, or to elderly
patients at high risk of hypoglycaemia37.
Meglitinides are usually used in combination with
metformin, a thiazolidinedione or insulin, although they
can be used as monotherapy. The results of RCTs have
 GLP-1RAs Meglitinides
Glucose
Sulfonylureas
GLUT1/2

GLP-1 receptor
Glucokinase

Glucose
cAMP PKA ATP K+
metabolism

Proinsulin
biosynthesis Mitochondrion KATP channel
(SUR1-Kir6.2)4
Piccolo Membrane
Rim2 Rab3A depolarization

EPAC2 Insulin Ca++

Insulin
granule
IP3 PLC
L-type Ca++
Rap1 channel
Ca++-/calmodulin-
dependent proteins
GPR40
PKC DAG agonists
Translocation

FFAR1 receptor
Priming

Docking

(KIWTG| Sulfonylureas, meglitinides and glucagon-like peptide 1 receptor agonists (GLP-1RAs) act on pancreatic
β|EGNNUVQKPETGCUGPWVTKGPVKPFWEGFKPUWNKPUGETGVKQP 6JGUGCIGPVUDKPFVQVJGE[VQUQNKEUWTHCEGQHVJGUWNHQP[NWTGC
TGEGRVQT|
574YJKEJKURCTVQHVJG#62UGPUKVKXG-KTRQVCUUKWOEJCPPGN$KPFKPIQHVJGUWNHQP[NWTGCQTOGINKVKPKFG
ENQUGUVJG-KTEJCPPGNRTGXGPVKPIRQVCUUKWOGHHNWZCPFVJGTGD[FGRQNCTK\KPIVJGRNCUOCOGODTCPG&GRQNCTK\CVKQP
QRGPUNQECNXQNVCIGFGRGPFGPVECNEKWOEJCPPGNUKPETGCUKPIVJGKPHNWZQHECNEKWOCPFCEVKXCVKPIECNEKWOFGRGPFGPV
UKIPCNNKPIRTQVGKPUVJCVEQPVTQNKPUWNKPGZQE[VQUKU).24#UGPJCPEGPWVTKGPVKPFWEGFKPUWNKPTGNGCUGOCKPN[XKCC
E#/2s'2#%OGFKCVGFRQVGPVKCVKQPQHITCPWNGGZQE[VQUKU&#)FKCE[NIN[EGTQN'2#%TCRIWCPKPGPWENGQVKFG
GZEJCPIGHCEVQT|).76INWEQUGVTCPURQTVGTKUQHQTO+2KPQUKVQNVTKURJQURJCVG4CR4CUTGNCVGFRTQVGKP|
2-#RTQVGKPMKPCUG|#2-%RTQVGKPMKPCUG|%2.%RJQURJQNKRCUG|%

shown that HbA1c reductions are similar to, or slightly less
than, those observed with sulfonylurea treatment when
meglitinides are used as monotherapy or as an add-on
to metformin37,83. Repaglinide can be used effectively in
conjunction with basal and biphasic insulins85,86. In an
RCT86 with treatment for 12 months, nonobese patients
with long-term T2DM (n = 102) were randomly assigned
to receive either repaglinide or metformin, both in com-
bination with biphasic insulin aspart 30/70 (30% soluble
insulin aspart and 70% intermediate-acting insulin
aspart), which was titrated to achieve an HbA1c level of
<6.5%. At the end of treatment, HbA1c reductions were
similar in both treatment groups (baseline versus study-
end HbA1c 8.15 ± 1.32 versus 6.72 ± 0.66% with metformin
and 8.07 ± 1.49% versus 6.90 ± 0.68% with repaglinide,
P = 0.2 for between-groups difference)86.
In a head-to-head RCT87 in which 150 drug-naive
patients were randomly assigned to receive either
repaglinide (0.5 mg per meal, maximum dose 4 mg
per meal) or nateglinide (60 mg per meal, maximum
dose 120 mg per meal) for 16 weeks, HbA1c reductions
from an average of 8.9% at baseline were greater with
repaglinide than nateglinide (−1.57% versus −1.04%,
P = 0.002). Reductions in FPG were also greater
with repaglinide than nateglinide (−57 mg/dl versus
−18 mg/dl, P <0.001)87.
Safety and adverse effects
Results of studies with repaglinide and nateglinide
have shown variable rates of hypoglycaemia, and gen-
erally less weight gain than with sulfonylureas83,88–92. In
a head-to-head RCT87, hypoglycaemia (blood glucose
<50 mg/dl) was more commonly associated with treat-
ment with repaglinide than with nateglinide (7% versus
0%). Weight gain was also slightly greater in the rep-
aglinide group (1.8 kg versus 0.7 kg)87. As co-treatments
with biphasic insulin, repaglinide and metformin
resulted in similar rates of hypoglycaemia, but weight
gain was less with metformin (difference in mean body
weight = −2.51 kg, 95% CI −4.07 kg to −0.95 kg)86.
 Meglitinides can bind to the sulfonylurea receptor 2
splice variants SUR2A and SUR2B, which are expressed
by cardiovascular tissues83,93. In the large NAVIGATOR
RCT94, nateglinide did not alter cardiovascular outcomes
in people with impaired glucose tolerance who either had,
or were at high risk of, cardiovascular disease. No asso-
ciation has been demonstrated between repaglinide and
either cardiovascular disease or cardiovascular risk65,83,95.
α-Glucosidase inhibitors (AGIs)
Acarbose was the first AGI to be introduced, in the
early 1990s; subsequently, miglitol and voglibose were
introduced in some countries. AGIs are widely used
in Asian populations that have diets in which complex
carbohydrates predominate37.
Mechanism of action
AGIs competitively inhibit α-glucosidase enzymes in
the brush border of enterocytes lining the intestinal
villi, preventing the enzymes from cleaving disaccha-
rides and oligosaccharides into monosaccharides37,96.
This action delays carbohydrate digestion and defers
absorption distally along the intestinal tract, reducing
blood-glucose excursions and lowering prandial insu-
lin levels37. Compared with controls, AGI treatment can
also increase postprandial GLP-1 secretion and reduce
secretion of glucose-dependent insulinotropic polypep-
tide (GIP)97,98. The affinities of AGIs vary for different
α-glucosidase enzymes, resulting in specific activity pro-
files (for example, acarbose has greater affinity for glyco-
amylase than for other glucosidases, whereas miglitol is
a stronger inhibitor of sucrase)37.
Pharmacokinetics
Acarbose is degraded by amylases and bacteria in the
small intestine; <2% of the unchanged drug is absorbed
(along with some of the intestinal degradation products).
Absorbed material is mostly eliminated in the urine
within 24 h37. Miglitol is almost completely absorbed,
and is eliminated unchanged in the urine37.
Pharmacodynamics
Typical HbA1c reductions with AGI treatment are ~0.5%,
mostly through reductions in postprandial glycaemia;
reductions depend upon the amount of complex carbo-
hydrate in the diet28. In a noninferiority RCT99 of Chinese
patients (n = 784) with newly diagnosed T2DM and mean
HbA1c of 7.5%, acarbose resulted in HbA1c reductions
similar to those with metformin (−1.1%, within groups
difference 0.01%, 95% CI −0.12% to 0.14%). However,
the sulfonylurea tolbutamide resulted in greater HbA1c
reductions compared with acarbose (−1.8% versus
−1.1%; mean difference 0.6%, 95% CI 0.2–1.0%) in
newly diagnosed drug-naive patients with T2DM (n = 96,
mean baseline HbA1c ~8%)100. Tolbutamide had a greater
effect on FPG than acarbose, whereas their effects on
postprandial glucose were similar 100.
Safety and adverse effects
Gastrointestinal adverse effects of AGIs (flatulence,
abdominal discomfort and diarrhoea) are commonly
encountered and can lead to treatment withdrawal.
Hypoglycaemia is uncommon. AGIs do not cause
weight gain, and they have no clinically significant drug
interactions.
The results of the STOP–NIDDM RCT showed that
acarbose reduces the risk of developing T2DM, delays
the onset of hypertension and reduces macrovascular
events by 49% compared with placebo, but the total
number of events was too small (n = 47) to draw firm
conclusions65,101,102. A large RCT103 assessing the impact
of acarbose on cardiovascular outcomes is ongoing.
Thiazolidinediones
Drugs derived from thiazolidinedione include piogl-
itazone, rosiglitazone and troglitazone. Troglitazone
was introduced in 1997 and withdrawn soon after
because of hepatotoxicity 28. Rosiglitazone and piogl-
itazone were introduced in 1999. Rosiglitazone was
discontinued in Europe and its use was restricted in
the USA in 2008 after reports of an association with
cardiovascular risk; the FDA lifted the restrictions in
2013. Pioglitazone was discontinued in 2011 in some
European countries pending enquires into a possible
risk of bladder cancer.
Mode of action
Thiazolidinediones are agonists of the peroxisome pro-
liferator-activated receptor gamma (PPAR-γ), a nuclear
receptor that is highly expressed in adipose tissue,
and to a lesser extent in muscle, liver, β cells, vascular
endothelium and macrophages37,104. PPAR-γ activation
alters gene expression, promoting adipogenesis, insulin
sensitivity and tissue glucose uptake, reducing inflam-
mation and altering energy balance104,105 in a tissue-
specific manner. PPAR-γ activation reduces hepatic
gluconeogenesis, modifies the blood lipid profile and
possibly improves β-cell viability 104,105. Differentiation
of pre-adipocytes into new small insulin-sensitive
adipocytes by PPAR-γ activation reduces circulating
levels of free fatty acids, which reduces ectopic
lipid accumulation in skeletal muscle and liver and
rebalances the Randle (glucose–fatty acid) cycle in
favour of glucose utilization by restricting availabil-
ity of free fatty acids as an energy source for hepatic
gluconeogenesis28.
Pharmacokinetics
Thiazolidinediones reach peak plasma levels within
1–2 h37. In the circulation, thiazolidinediones are almost
entirely bound to plasma proteins, but their concentra-
tions are not sufficient to interfere with other protein-
bound drugs 37 . Pioglitazone is metabolized by
cytochrome P450 2C8 (CYP2C8) and CYP3A4 to weakly
active metabolites that are eliminated via bile, whereas
rosiglitazone is metabolized by CYP2C9 and CYP2C8
to inactive metabolites that are excreted via urine37,106.
Rifampicin induces expression of CYP3A4, resulting in
a reduction in levels of rosiglitazone and pioglitazone,
whereas the lipid-lowering fibrate gemfibrozil inhibits
CYP2C8, leading to accumulation of rosiglitazone and
pioglitazone106.
Pharmacodynamics
Maximal doses of thiazolidinediones can reduce HbA1c
by 0.7–1.6% when used as monotherapy or in combina-
tion with metformin, sulfonylureas or insulin104,107. In an
RCT108, patients with T2DM receiving metformin (n = 630,
mean age ~56 years, mean diabetes duration ~5.5 years,
baseline mean HbA1c 8.5–8.7%) were randomly assigned
to either pioglitazone or gliclazide as an add-on treat-
ment. After 2 years, the changes in HbA1c were similar
in the two arms (−0.89% with pioglitazone and −0.77%
with gliclazide, P = 0.2 for between-groups difference),
whereas pioglitazone resulted in greater reductions
in FPG (−1.8 mmol/l versus −1.1 mmol/l, P <0.001)108. In
another RCT108, patients with T2DM receiving a sulfon-
ylurea (n = 639, mean age ~60 years, mean T2DM dura-
tion ~7 years, baseline mean HbA1c 8.8%) were randomly
assigned to either pioglitazone or metformin as an add-on
treatment. After 2 years, the changes in HbA1c were sim-
ilar in the two arms (−1.03% with pioglitazone versus
−1.16% with gliclazide, P = 0.17 for between-groups dif-
ference); reductions in FPG (around ~2 mmol/l) were also
similar in the two arms108. Onset of the glucose-lowering
effect of thiazolidinediones is gradual, taking 2–3 months
to reach maximum effect37. The ADOPT trial15, in which
4,360 patients with T2DM (mean age 56–58 years, base-
line HbA1c 7.4%, mostly <2 years T2DM duration) were
randomly assigned to glyburide, metformin or rosigli-
tazone, showed that rosiglitazone as monotherapy has a
more prolonged effect on glycaemic control (measured
by HbA1c and FPG) than metformin or glyburide over
5 years. The glucose-lowering efficacy of thiazolidine-
diones is generally gradual in onset over several weeks,
varies considerably between individuals, and no defi-
nite predictors are known to identify responders versus
nonresponders29.
Safety and adverse effects
Thiazolidinediones do not affect the risk of hypoglycae-
mia when used as monotherapy or in combination with
metformin. Oedema (often identified by rapid weight
gain) has been reported in 4–6% of patients receiving
thiazolidinediones104; the observed fluid retention is the
result of renal sodium reabsorption mediated by increased
expression of sodium channel proteins in collecting duct
epithelium28. Thiazolidinediones are associated with
weight gain of 2–3 kg for each 1% drop in HbA1c, whether
used as monotherapy or in combination with metformin
or insulin104. The weight gain is usually in subcutaneous
adipose tissue, whereas visceral fat is either reduced or
unaltered104,109. In the ADOPT trial15, the weight gain with
rosiglitazone over 5 years was greater than with gliben-
clamide (glyburide; treatment difference 2.5 kg, 95% CI
2.0–3.1 kg, P <0.001), whereas no difference was observed
in waist circumference (treatment difference 0.77 cm,
95% CI −0.21 cm to 1.76 cm, P = 0.12).
The results of RCTs and observational studies show
that, compared with control groups, long-term treat-
ment with thiazolidinediones lowers bone density and
doubles the risk of fractures in patients with T2DM, par-
ticularly in women110. In the ACCORD trial111, women
who received a thiazolidinedione had double the risk
of nonspinal fracture compared with those not using a
thiazolidinedione; this risk was reduced after discontin-
uation of the thiazolidinedione. A meta-analysis of RCTs
showed that, compared with metformin, sulfonylureas
or placebo, thiazolidinediones reduce bone mineral den-
sity at the lumbar spine (difference −1.1%, 95% CI −1.6%
to −0.7%, P <0.0001), total hip (−1.0%, 95% CI −1.4% to
−0.6%, P <0.0001), forearm (−0.9%, 95%  CI −1.6%
to −0.3%, P = 0.007) and femoral neck (−0.7%, 95% CI
−1.4% to 0.0%, P = 0.06), effects that were not reversed
after 1 year of stopping treatment in some studies112.
The cardiovascular safety of thiazolidinediones was
questioned in a controversial meta-analysis that showed
increased adverse cardiovascular outcomes in patients
treated with rosiglitazone compared with controls with-
out rosiglitazone, prompting withdrawal of rosiglitazone
in Europe and restricted use in the USA65,113. However,
when the FDA re-examined the data from the RECORD
study, no significant effect on cardiovascular risk
was found65,114.
Pioglitazone is a ligand for PPAR-α (as well as PPAR-γ),
and through PPAR-α it mitigates cardiovascular risk by
positive effects on plasma levels of HDL cholesterol, reduc-
tions in plasma triglycerides and small dense LDL choles-
terol particles, with the production of larger, more buoyant
particles115. Thiazolidinediones can also have beneficial
effects on blood pressure and endothelial function65, but
compared with pioglitazone, rosiglitazone increases levels
of plasma LDL cholesterol and triglycerides65.
In the PROACTIVE trial32, compared with placebo,
pioglitazone was associated with a numerical but non-
significant reduction in the composite outcome of all-cause
mortality, nonfatal myocardial infarction, stroke, acute
coronary syndrome, endovascular or surgical intervention
in the coronary or leg arteries and amputation above the
ankle (HR 0.90, 95% CI 0.80–1.02, P = 0.095). However,
pioglitazone significantly lowered the occurrence of the
secondary end point, a composite of all-cause mortal-
ity, nonfatal myocardial infarction and stroke (HR 0.84,
95% CI 0.72–0.98, P = 0.027)32. In addition, pioglitazone
reduced the risks of subsequent myocardial infarction
and recurrent stroke by 16% and 47%, respectively 65,116,117.
Nonetheless, the risk of heart failure was higher in
the pioglitazone group than the placebo group in the
PROACTIVE trial, although this risk was not associated
with increased mortality 65.
However, both rosiglitazone and pioglitazone can
cause congestive heart failure in patients who already
have diastolic dysfunction, because of the propensity
for oedema65. The effects of rosiglitazone on coronary
artery disease are not clear, but pioglitazone might be
beneficial65,118–122.
DPP-4 inhibitors
The currently available DPP-4 inhibitors (sitagliptin,
vildagliptin, saxagliptin, linagliptin and alogliptin)123
are licensed as monotherapy, dual therapy, triple ther-
apy and in combination with insulin, but there are some
minor variations in licensing between agents. In addi-
tion, once-weekly DPP-4 inhibitors (omarigliptin and
trelagliptin) are licensed in Japan124,125.
Mechanism of action
The action of DPP-4 inhibitors causes elevation of cir-
culating levels of incretin hormones, notably GLP-1
and GIP. The incretin effect is the ability of intestinal
factors to enhance nutrient-induced insulin responses
during feeding by 50–70% in healthy individuals126,127;
this effect is much diminished in T2DM. GIP is secreted
by K cells in the duodenum and jejunum in response to
ingestion of carbohydrates and lipids128–130. In addition
to its incretin effect, GIP reduces gastric acid secretion
and has roles in adipogenesis and possibly β-cell prolif-
eration128,130–133. GLP-1 is secreted by L cells mainly in
the distal ileum and colon128,130, and accounts for most
of the incretin effect 128,134, including insulin biosyn-
thesis135,136. Additionally, GLP-1 causes a reduction in
glucagon secretion, and has extrapancreatic actions
that enhance satiety and delay gastric emptying (see
Supplementary information S2 (box))127,134,137–139.
GIP and GLP-1 are rapidly degraded by DPP-4
(REF. 128), which acts on peptides to cleave N-terminal
dipeptides with alanine (as in the incretins) or proline at
position N2 (REF. 130). DPP-4 exists free in the circulation
and also attached to endothelial cells130,140, and is widely
expressed in human tissues, including the intestine and
portal system130. GLP-1 and GIP are generally inacti-
vated almost immediately following secretion, and have
half-lives of <2 min and 5–7 min, respectively 128,130,141,142.
Compared with placebo treatment, DPP-4 inhibition
results in a 2–3-fold increase in postprandial active GLP-1
levels143,144. Unlike GLP-1 receptor agonists (GLP-1RAs),
which can have an effect equivalent to a >10-fold increase
in GLP-1, DPP-4 inhibitors do not delay gastric emptying
or increase satiety and weight loss, but they do avoid
initial nausea and vomiting 145,146.
Pharmacokinetics
Currently available DPP-4 inhibitors can produce
77–99% inhibition of DPP-4 activity and are appropri-
ate for once-daily dosing, except for vildagliptin (twice-
daily), and omarigliptin and trelagliptin (once-weekly).
They are predominantly excreted in the urine, except for
linagliptin, which does not require dose adjustment in
patients with CKD (see Supplementary information S3
(table))123,147–151.
DPP-4 inhibitors have little or no interaction with
other glucose-lowering agents or drugs commonly used in
patients with T2DM123,152, possibly because DPP-4 inhib-
itors are neither inducers nor inhibitors of cytochrome
P450 isoforms, and are not appreciably bound to plasma
proteins152. However, saxagliptin is metabolized to an
active metabolite by CYP3A4 and CYP3A5 (REFS 123,152).
Pharmacodynamics
On average, DPP-4 inhibitors reduce postprandial glucose
excursions by ~3 mmol/l, and FPG by ~1.0–1.5 mmol/l28,123.
A meta-analysis153 that assessed the efficacy of DPP-4
inhibitors as monotherapy or as add-on therapy to
other oral agents included placebo-controlled or active-
controlled RCTs of DPP-4 inhibitors (n = 98 trials, 24,163
patients) of 12–54 weeks duration, with ≥30 patients in
each treatment arm. The mean ages of the participants
in all but two of these studies were 50–62 years; 88 of the
98 trials included were double-blinded and 10 were open-
label design153. The results showed that DPP-4 inhibitors
reduce HbA1c by −0.77% (95% CI −0.82% to −0.72%)
from an average baseline of 8.05%153. In 18 RCTs with
a duration of 52–54 weeks, DPP-4 inhibitors resulted in
HbA1c reductions of −0.84% (95% CI −0.99% to −0.68%,
P <0.0001), whereas in 26 RCTs of 12–18 weeks dura-
tion, the HbA1c reduction was −0.68% (95% CI −0.75%
to −0.61%, P <0.0001)153.The HbA1c reductions were
largely similar across the class, but results from direct
head-to-head trials are limited. In this meta-analysis, the
HbA1c reductions according to DPP-4 inhibitor were:
vildagliptin 50 mg (n = 26, age 56.3 years, baseline HbA1c
8.06%) −0.88% (95% CI −1.00% to −0.75%, P <0.0001);
sitagliptin 100 mg (n = 37, age 55.2 years, baseline HbA1c
8.05%) −0.79% (95% CI −0.87% to −0.71%, P <0.0001);
saxagliptin 5 mg (n = 13, age 55.4 years, baseline HbA1c
8.01%) −0.70% (95% CI −0.79% to −0.62%, P <0.0001);
linagliptin 5 mg (n = 13, age 59.0 years, baseline HbA1c
8.05%) −0.55% (95% CI −0.65% to −0.45%, P <0.0001);
alogliptin 25 mg (n = 11, age 55.2 years, baseline HbA1c
8.14%) −0.76% (95% CI −0.86% to −0.66%, P <0.0001)153.
Reductions in HbA1c were greater in patients with base-
line HbA1c >9.0%, compared with HbA1c ≤9.0%153. For
RCTs with basal HbA1c <7.5% (n = 8, age 57.4 years,
baseline HbA 1c 7.32%) the HbA 1c reduction was
−0.63% (95% CI −0.78% to −0.48%, P <0.0001); for
basal HbA1c 7.5–8.0% (n = 28, age 57.6 years, baseline
HbA1c 7.82%) the reduction was −0.70% (95% CI −0.76
to −0.63, P <0.0001); for basal HbA1c 8.0–8.5% (n = 34,
age 55.9 years, baseline HbA1c 8.15%) the reduction
was −0.72% (95% CI −0.79% to −0.64%, P <0.0001); for
basal HbA1c >9.0% (n = 30, age 54.2 years, baseline HbA1c
8.63%) the reduction was −0.93% (95% CI −1.02% to
−0.84%, P <0.0001)153.
A meta-analysis of 27 reports of 19 studies, including
7,136 patients, showed that DPP-4 inhibitor monotherapy
was associated with a smaller decline in HbA1c than met-
formin monotherapy (weighted mean difference (WMD)
0.20%, 95% CI 0.08–0.32%, Poverall effect = 0.001). DPP-4
inhibitors combined with metformin produced smaller
declines than GLP-1RAs combined with metformin
(WMD 0.49%, 95% CI 0.31–0.67%, Poverall effect <0.001) and
sulfonylureas combined with metformin (WMD 0.07%,
95% CI 0.03–0.11%, Poverall effect <0.001), but similar declines
to pioglitazone combined with metformin (WMD 0.09%,
95% CI −0.07% to 0.24%, Poverall effect = 0.28)154. In the
studies comparing monotherapies, trial durations were
24–206 weeks, and participants had mean T2DM dura-
tion of 1.0–4.4 years and mean HbA1c of 7.2–9.6%. In the
trials of combination therapies, mean T2DM duration was
5.0–7.3 years and mean HbA1c was 7.3–8.5%154.
The comparison of the efficacy of DPP-4 inhibi-
tors with sulfonylureas is complicated by multiple fac-
tors including study duration, renal function and the
choice of sulfonylurea for the active comparator 155. In
a meta-analysis155 of 12 RCTs of ≥18 weeks duration
that compared sulfonylureas head-to-head with DPP-4
inhibitors, the mean changes from baseline in HbA1c were
modestly but significantly smaller with DPP-4 inhibitors
than with sulfonylureas (difference of mean changes in
HbA1c 0.105, 95% CI 0.103–0.107, P <0.0001). However,
several RCTs of 1–3 years duration showed similar HbA1c
reductions for DPP-4 inhibitors and sulfonylureas150,155–164.
The glucose-lowering efficacy of DPP-4 inhibitors is
greater in Asian patients with T2DM than in other ethnic
groups (between-group HbA1c difference −0.26%, 95% CI
−0.36% to−0.17%, P <0.001), and might be affected by
genetic factors such as a TCF7L2 gene variant 165,166. A
meta-analysis of RCTs of ≥76 weeks duration suggested
that the effect of DPP-4 inhibitors is not durable and
lessens during the second year of treatment167.
Head-to-head comparisons of DPP-4 inhibitors
The number of head-to-head trials comparing DPP-4
inhibitors is limited. The results of an RCT168 compar-
ing saxagliptin and sitagliptin as add-on treatment to
metformin in 810 patients (age 58.4 years, diabetes
duration 6.3 years, baseline HbA1c 7.7%) showed that
HbA1c reductions over 18 weeks are similar with both
treatments (adjusted mean changes in HbA1c −0.52 and
−0.62%, respectively; between-group difference 0.09%,
95% CI −0.01% to 0.20%). However, sitagliptin resulted
in a slightly greater reduction in FPG (−0.60 mmol/l for
saxagliptin versus −0.90 mmol/l for sitagliptin; treatment
difference 0.30 mmol/l, 95% CI 0.08–0.53 mmol/l)168.
In another RCT169, 148 patients with T2DM and
estimated glomerular filtration rate (eGFR) <30 ml/
min/1.73m2, who were either drug-naive or treated with
any glucose-lowering agents were randomly allocated to
vildagliptin 50 mg or sitagliptin 25 mg once daily. The
treatments resulted in similar reductions in HbA1c over
24 weeks (adjusted mean change in HbA1c −0.54% from a
baseline of 7.52% with vildagliptin versus −0.56% from a
baseline of 7.80% with sitagliptin, P = 0.874). Vildagliptin
lowered FPG by 0.47 ± 0.37 mmol/l, whereas FPG
increased in the sitagliptin group by 0.16 ± 0.43 mmol/l,
but the difference between groups was not significant
(P = 0.185)169.
In a phase III, noninferiority RCT170, 243 patients
with T2DM that was inadequately controlled by diet
and exercise were randomly assigned to receive trelaglip-
tin (100 mg once weekly), alogliptin (25 mg daily) or
placebo for 24 weeks. Trelagliptin was noninferior to
alogliptin and resulted in similar reductions in HbA1c
(−0.33% versus −0.45%, respectively; least-squares mean
difference (LSMD) 0.11%, 95% CI −0.054% to 0.281%).
Both trelagliptin and alogliptin significantly reduced
mean HbA1c compared with placebo (P <0.0001)170.
In another RCT171, 412 patients with T2DM, who
were drug-naive or on oral glucose-lowering treatments,
were randomly assigned to receive omarigliptin 25 mg
weekly, sitagliptin 50 mg daily or placebo, for 24 weeks.
At baseline, patients had mean HbA1c of 7.9%, 8.0% and
8.1%, respectively, in the omarigliptin, sitagliptin and
placebo groups171. Omarigliptin treatment resulted in
HbA1c reduction of −0.66% (95% CI −0.76% to −0.57%),
which was significantly greater than with placebo
(P <0.001) and similar to sitagliptin (LSMD −0.02%,
95% CI −0.15% to 0.12%), meeting the prespecified
noninferiority criterion171.
Safety and adverse effects
DPP-4 inhibitors are generally well tolerated, and the
incidence of adverse effects is similar to placebo and
lower than other glucose-lowering agents154,172. The inci-
dence of gastrointestinal symptoms is lower with DPP-4
inhibitors than with metformin or a GLP-1RA154. The
risk of hypoglycaemia in patients treated with a DPP-4
inhibitor is very low except when combined with
sulfonylureas or insulin123,154,172.
DPP-4 has many substrates other than incretins,
including bradykinin, encephalins, neuropeptide Y,
peptide YY1–36, gastrin-releasing polypeptide, sub-
stance P, insulin-like growth factor 1, vasostatin 1, the
α chains of thyrotropin, luteinizing hormone, chorionic
gonadotropin and several chemokines, such as C-C motif
chemokine 2 (monocyte chemotactic protein 1)173; how-
ever, no adverse effects relating to these substrates have
been observed in clinical trials28,123,148. In addition, DPP-4
is the T-cell activation antigen CD26, but no untoward
immune-related effects have been demonstrated, either in
Dpp4 knockout mice or with the use of DPP-4 inhibitors
in animals or humans28.
Several meta-analyses and pooled analyses have
shown that DPP-4 inhibitors (individually and as a
class) are associated with reductions in cardiovascular
events65,174. However, these studies were retrospective
and not specifically designed to examine the effect of
DPP-4 inhibitors on the incidence of cardiovascular
disease65. The results of three RCTs (SAVOR–TIMI175,176,
EXAMINE177 and TECOS178) demonstrated that saxa-
gliptin, alogliptin and sitagliptin are not associated with
the risk of adverse cardiovascular outcomes65,175–178. The
populations of patients in these trials were each slightly
different. SAVOR–TIMI175,176 included patients with
T2DM with either a history of, or a risk of, cardiovascu-
lar events. EXAMINE177 included patients with T2DM
and an acute myocardial infarction or hospitalization for
unstable angina in the preceding 15–90 days. TECOS178
included patients with T2DM who were >50 years old
and had established cardiovascular disease.
These studies were designed to look specifically at
the effects of DPP-4 inhibitors on cardiovascular safety
so that patients in the placebo arms received other glu-
cose-lowering therapies to minimize any differences in
HbA1c between arms. In the SAVOR–TIMI study 175,176,
saxagliptin treatment was associated with a 3.5% inci-
dence of hospitalization for heart failure, compared with
2.8% in the placebo arm (P = 0.007), with no increase in
mortality, and this difference was independent of base-
line renal function, although compared with placebo,
saxagliptin reduced the development and progression of
microalbuminuria65,175,176. No effect on heart failure was
observed in the EXAMINE177 or TECOS178 trials, and
the mechanism underlying the effect that was seen with
saxagliptin is unclear. An ongoing study (CAROLINA80)
has been designed to examine the effect of linagliptin
on cardiovascular outcomes with an active comparator
(glimepiride) rather than placebo.
The individual results of the SAVOR–TIMI, EXAMINE
and TECOS trials did not show any increased risk of pan-
creatitis or pancreatic cancer175–179, but a meta-analysis of
these RCTs did demonstrate a significantly increased risk
of acute pancreatitis in patients using DPP-4 inhibitors
compared with those receiving standard care (OR 1.82,
95% CI 1.17–2.82, P = 0.008)180.
GLP-1RAs
Exenatide (twice daily) was the first GLP-1RA, and
was introduced in 2005. Two once daily GLP-1RAs
(liraglutide and lixisenatide) and three once weekly
GLP-1RAs (exenatide, albiglutide and dulaglutide)
are also now available for combination therapy with
oral glucose-lowering agents and basal insulin (except
exenatide once weekly, which is not licensed to be used
with basal insulin). Dulaglutide and albiglutide are also
licensed as monotherapy in patients who are intolerant
to metformin.
Exenatide (synthetic exendin-4), a peptide originally
isolated from saliva of the lizard Heloderma suspectum
(Gila monster)128,181, has 53% homology with human
GLP-1 and contains an Ala8Gly substitution that con-
fers resistance to degradation by DPP-4 (REFS 128,182).
Exenatide once weekly sustained-release formulation
consists of exenatide embedded within biodegradable
polymeric microspheres of poly(DL-lactic-co-glycolic
acid)183. Liraglutide is a true analogue of GLP-1 with
the addition of a 16-carbon fatty acid chain attaching
Lys26 to albumin, to mask the DPP-4 cleavage site184.
Albiglutide has two copies of GLP-1 in series, each with
an Ala8Gly substitution, and this molecule is fused to
albumin185. Lixisenatide is an exendin-4 analogue with
six Lys residues added at the C terminus to confer resist-
ance to DPP-4 (REF. 186). Dulaglutide has two copies of a
GLP-1 analogue (with amino acid substitutions Ala8Gly,
Gly22Glu and Arg36Gly) covalently linked to an
Fc fragment of human IgG4 (REF. 187).
Mechanism of action
GLP-1RAs mimic GLP-1 and activate the GLP-1 recep-
tor, potentiating nutrient-induced insulin secretion
(FIG. 3), contributing to reductions in fasting glycaemia
and postprandial glycaemia, and to weight loss188 (see
Supplementary information S2,S4,S5 (box, table, table)).
Therapeutic concentrations of GLP-1RAs are far higher
than physiological levels of GLP-1, and although GLP-1
deficiency has been described in patients with T2DM, this
deficiency is not a universal characteristic of the disease188.
Pharmacokinetics
GLP-1RAs are delivered by subcutaneous injection.
Exenatide is rapidly absorbed189. Tmax is ~2 h, half-life
is 3–4 h189 and elimination is mostly renal by glomer-
ular filtration and proteolytic degradation190–192 (see
Supplementary information  S6 (table)). Relative to
patients with normal renal function, exenatide clearance is
decreased by 36% in patients with moderate renal disease
(in whom it should be used with caution) and by 84% in
those with severe renal disease (in whom it should not be
used)193. The once weekly exenatide reaches therapeutic
levels within 2 weeks and maximum concentrations by
6 weeks194. The half-life of liraglutide is 10–15 h, with a
Tmax of 9–12 h195–197. Lixisenatide has a half-life of 2–4 h
and a Tmax of 1–2 h198, and exerts its main effect on the
meal immediately after injection. Albiglutide has a Tmax
of 3–5 days and a half-life of 6–7 days199. Dulaglutide
has a Tmax of 12–72 h, a half-life of ~4 days and reaches
steady-state levels by 2 weeks200 (see Supplementary infor-
mation S6 (table)). GLP-1RAs are not recommended in
severe renal disease; they have limited drug interactions,
but can affect the availability of other medicines, such
as acetaminophen (paracetamol) and statins because of
the delay in gastric emptying (except for exenatide once
weekly, which has a minor effect on gastric emptying)27,201.
Pharmacodynamics
The efficacy of GLP-1RAs has been explored in large
programmes of placebo-controlled and active-compar-
ator RCTs202–238 (see Supplementary information S4,S5
(tables)).
Effect on glycaemic measures. Exenatide (twice daily)
significantly reduces measures of glycaemic con-
trol when used as monotherapy or add-on therapy
(see Supplementary information S4 (table))239–243. A
meta-analysis244 of RCTs in which exenatide was an
add-on to existing metformin therapy for 16–30 weeks
showed that exenatide lowers HbA1c by 0.8% from an
average baseline of 8.1 ± 0.6%. The effect of exenatide
on HbA1c reduction was greater in patients with baseline
HbA1c >9% than in those with HbA1c ≤9%239, and was
maintained at 3 years240 and only deteriorated modestly
through 6 years245,246.
Liraglutide improves glycaemic control when used
as monotherapy or add-on therapy 239,241,247,248 (see
Supplementary information S4 (table)). Compared with
glimepiride 8 mg daily, liraglutide 1.2–1.8 mg daily mon-
otherapy resulted in greater reductions in HbA1c from an
average baseline of 8.3% (glimepiride −0.6%, liraglutide
1.2 mg −0.9% and liraglutide 1.8 mg −1.1%; treatment
difference for liraglutide 1.2 mg −0.31%, 95% CI −0.54%
to −0.08%, P = 0.008; treatment difference for liraglutide
1.8 mg −0.60%, 95% CI −0.83% to −0.38%, P <0.0001).
A similar effect was seen for reductions in FPG (treat-
ment difference for liraglutide 1.2 mg −0.63 mmol/l,
95% CI −1.17 mmol/l to −0.09 mmol/l, P = 0.02; treat-
ment difference for liraglutide 1.8 mg −0.99 mmol/l,
95% CI −1.53 mmol/l to −0.45 mmol/l, P <0.001) and
also postprandial glucose over 104 weeks247. In pooled
patient data from seven phase III RCTs from the lira-
glutide development programme, with 26  weeks
of liraglutide 1.8 mg, HbA1c reductions were less in
patients with baseline HbA1c ≤7.5% than in those with
HbA1c >9.0% (−0.7% versus −1.8%)249.
Lixisenatide significantly decreases HbA1c and post-
prandial glucose when used as monotherapy or add-on
therapy 210–215,250–257. In a meta-analysis of RCTs, com-
pared with placebo, lixisenatide treatment produced
reductions in 2 h postprandial glucose from baseline
(LSMD −4.9 mmol/l, P <0.001), glucose excursion
(LSMD −4.5 mmol/l, P <0.001) and postprandial gluca-
gon (LSMD −19.0 ng/l, P <0.001)256. Compared with pla-
cebo, lixisenatide also reduced HbA1c and postprandial
glucose, but not FPG, when added to basal insulin257.
 Exenatide once weekly reduces HbA1c, FPG and post-
prandial glucose when used as monotherapy or add-on
treatment218,239,241,258. Exenatide once weekly monotherapy
has been noninferior to metformin, superior to sitagliptin
and similar to pioglitazone with regard to HbA1c reduc-
tion in RCTs at 26 weeks239,258. Addition of exenatide once
weekly to metformin is more effective for the achieve-
ment of glucose control than addition of either sitagliptin
or pioglitazone to metformin218,239. In a study involving
456 patients with T2DM treated with metformin alone
or with a sulfonylurea, addition of exenatide once weekly
resulted in similar HbA1c reductions to addition of insu-
lin glargine; the effect of exenatide once weekly persisted
at 3 years222,236,239. Similarly, addition of exenatide once
weekly to oral glucose-lowering medication resulted in
greater HbA1c reductions over 26 weeks than addition
of once daily or twice daily insulin detemir 239,259. In the
extension phase of the DURATION-1 trial260, patients
received exenatide once weekly for up to 5 years, and
improvements in HbA1c and FPG were maintained over
this period. However, 40% of patients did not complete
the study. Most of the loss of follow-up was because of
withdrawal of consent, and only eight patients withdrew
because of “loss of glucose control”. No differences were
identified in baseline characteristics between those who
completed and did not complete the study, and HbA1c
reduction at 5 years was evident in the intention-to-treat
analysis (−1.2% ± 0.1%) and the analysis of patients who
completed the extension (−1.6% ± 0.1%).
Albiglutide has beneficial effects on glycaemic con-
trol when used as monotherapy or add-on therapy in
phase III studies250,261,262. In an RCT lasting 104 weeks,
treatments were added to metformin, and albiglutide
provided significantly greater reductions in HbA1c and
FPG than placebo, sitagliptin or glimepiride226. As an
add-on to treatment with metformin and sulfonylurea,
albiglutide did not meet the prespecified noninferior-
ity margin for the difference in the change of HbA1c of
0.3% compared with pioglitazone over 52 weeks223. As
an add-on to metformin (with or without sulfonylurea),
albiglutide resulted in similar HbA1c reductions to insu-
lin glargine over 52 weeks224. As an add-on to insulin
glargine, albiglutide was noninferior to insulin lispro at
26 weeks, but did not meet the noninferiority margin
at 52 weeks250,263.
Dulaglutide 0.75 mg and 1.5 mg weekly treatments
were more effective than metformin and sitagliptin when
used as monotherapy or as add-on therapy to other oral
glucose-lowering treatments over 52 weeks232,234,250. In
addition to metformin and sulfonylureas over 52 weeks,
compared with daily insulin glargine, dulaglutide 1.5 mg
weekly was more effective and dulaglutide 0.75 mg was
noninferior for the reduction of HbA1c from baseline237.
A meta-analysis of RCTs of ≥12 weeks duration in
which information about ethnicity was available showed
that the WMD in HbA1c for GLP-1RA treatment com-
pared with placebo was −1.16% (95% CI −1.48% to
−0.85%) in the pool of studies involving ≥50% Asian
participants, and −0.83% (95% CI −0.97% to −0.70%) in
the studies with <50% Asian participants (between-group
difference −0.32%, 95% CI −0.64% to −0.01%, P = 0.04)264.
Effect on weight. GLP-1RAs are associated with reduc-
tions in body weight and waist circumference, but with
much variation in individual responses and within-class
differences250,265–268 (see Supplementary information S5
(table)). In a meta-analysis266 of 15 RCTs, the combi-
nation of GLP-1RAs with basal insulin was shown to
result in mean weight loss of −3.22 kg (95% CI −4.90 kg
to −1.54 kg).
Effect on blood pressure. Results from individual RCTs
and meta-analyses have shown that GLP-1RAs result
in modest, but significant, reductions in systolic blood
pressure, compared with placebo or insulin269–271 (see
Supplementary information S5 (table)). This effect is inde-
pendent of baseline blood pressure and the influence of
the GLP-1RA on HbA1c or body weight269. A reduction in
diastolic blood pressure compared with placebo has also
been observed with exenatide twice daily (−1.08 mmHg,
95% CI −1.78 mmHg to −0.33 mmHg)270.
Other effects. A meta-analysis has shown that GLP-1RAs
modestly reduce levels of total cholesterol, LDL choles-
terol and triglycerides, but do not improve HDL cho-
lesterol levels, in comparison with placebo or active
comparators272.
Safety and adverse effects
GLP-1RAs are generally well tolerated; the most com-
mon adverse effect is nausea, which is usually transient,
resolving over 4–8 weeks, and which can be minimized
by progressively increasing the dose27,28,239,250. The risk
of hypoglycaemia in patients receiving GLP-1RAs is
low, unless they are combined with insulin or sulfonyl-
ureas27,28,239,250. Injection-site reactions are common
with some GLP-1RAs, including exenatide once weekly
(≤17.6%) and albiglutide (≤22%)250. The occurrence of
antibodies is also commonly associated with GLP-1RA
treatment, but with little apparent clinical relevance, and
generally with no influence on glycaemic control, except
very occasionally in patients receiving exenatide once
weekly who have high antibody titres28,194,239,250.
The possible association between GLP-1RAs and the
risk of pancreatitis and pancreatic cancer has received
much attention, but to date no definite causal link has
been found273. Meta-analyses have shown no significant
increase in acute pancreatitis with GLP-1RA treatment
in patients with T2DM179,274,275. In addition, results from
cardiovascular safety trials have not shown a significant
increase in pancreatitis with GLP-1RAs276. The recom-
mendation in the product labelling for GLP-1RAs to
avoid therapy in patients with a history of pancreatitis
and to discontinue treatment if pancreatitis develops is
considered appropriate28. Thyroid C-cell hyperplasia and
medullary cell carcinoma were also raised as possible
concerns in preclinical studies in rodents, but clinical
studies have not identified any substantial problems with
GLP-1RA treatment 28,239,250.
The results of preclinical studies showed that
GLP-1RAs have cardioprotective effects in heart fail-
ure and following myocardial ischaemia65. GLP-1RAs
can have favourable effects on many cardiovascular risk
factors, such as weight loss, blood pressure, endothelial
function, inflammation, plasminogen activator inhib-
itor-1, postprandial lipaemia and LDL cholesterol65.
Results of studies in patients with and without T2DM
have shown a beneficial effect of GLP-1RAs on left
ventricular function in patients with heart failure and
on myocardial function and the myocardial salvage
index following ischaemia65,277. However, GLP-1RAs
often stimulate the resting heart rate by ~3 bpm, most
likely by activating GLP-1R in the sinoatrial node65. In
an RCT271 with 24-h ambulatory heart-rate monitor-
ing, dulaglutide 1.5 mg was associated with increased
heart rate compared with placebo (LSMD 2.8 bpm,
95% CI 1.5–4.2 bpm), unlike dulaglutide 0.75 mg and
exenatide271,278. Large RCTs assessing the cardiovascular
safety of liraglutide, semaglutide, exenatide once weekly
and dulaglutide are ongoing 65. No adverse cardiovascular
outcomes have been reported in patients with T2DM and
established cardiovascular disease who were treated with
lixisenatide276.
Head-to-head comparisons of GLP-1RAs
As several GLP-1RAs are available, with different chem-
ical structures and formulations, the different phar-
macokinetic and pharmacodynamic profiles seen in
head-to-head trials could influence clinical decision-
making 205,216,219–221,227,230,231,238,279 (TABLE 2). Overall, liraglu-
tide 1.8 mg and dulaglutide 1.5 mg seem to have the great-
est effects on HbA1c, and liraglutide 1.8 mg and exenatide
once weekly have the largest effect on weight reduction.
Albiglutide has less effect on HbA1c and weight reduction
than other GLP-1RAs, but is associated with fewer gastro-
intestinal adverse effects. Once weekly preparations are
associated more with injection-site reactions than once
daily or twice daily agents.
In general, longer-acting GLP-1RAs produce greater
reductions in FPG, but have less effect on postpran-
dial glucose excursions, compared with shorter-acting
GLP-1RAs280,281. The effect on postprandial glucose is
at least partly mediated by delayed gastric emptying,
and occurs more with short-acting GLP-1RAs than
with long-acting GLP-1RAs, which are subject to tachy-
phylaxis brought on by chronic elevation of plasma
GLP-1 (REF. 280). In addition, lixisenatide, in contrast to
liraglutide, strongly suppresses postprandial glucagon
secretion280. Patient satisfaction is greater in those receiv-
ing exenatide once weekly or liraglutide, compared with
exenatide twice daily 279.
GLP-1RAs versus insulin
In a meta-analysis282 of RCTs that compared GLP-1RAs
with basal insulin progressively titrated to achieve FPG
targets in patients with T2DM, GLP-1RAs resulted in
greater reductions in HbA1c (mean net change −0.14%,
95% CI −0.27% to −0.02%, P = 0.03) and weight (−4.40 kg,
95% CI −5.23 kg to −3.56 kg, P <0.01), but less reduction
in FPG (1.18 mmol/l, 95% CI 0.43–1.93 mmol/l, P <0.01).
GLP-1RAs were also associated with greater reduc-
tions in postprandial glucose compared with insulin282.
Hypoglycaemia was reported to reduce in the GLP-1RA
group (HR 0.45, 95% CI 0.27–0.76, P <0.01)282. In a
separate RCT283, dulaglutide resulted in greater reduc-
tion in HbA1c than insulin glargine, when added to
insulin lispro.
Insulin–GLP-1RA combination
To simplify the co-administration of basal insulin and
GLP-1RAs, these two agents have been combined into
a single injection, a fixed-ratio combination (IDegLira),
which was launched in the UK in 2014 (REF.  147) .
IDegLira combines 50 U of insulin degludec with 1.8 mg
of liraglutide147; the combination is titrated in the same
way as insulin alone147.
In a 26-week RCT284 involving insulin-naive patients,
HbA1c decreased by 1.9% ± 1.1% with IDegLira, com-
pared with 1.4% ± 1.0% with insulin degludec and
1.3% ± 1.1% with liraglutide. The IDegLira group had
fewer reports of nausea than the liraglutide group and
lower incidence of hypoglycaemia than the insulin deglu-
dec group284. These benefits were maintained at 52 weeks,
with HbA1c reductions of 1.84%, 1.40% and 1.21% for
IDegLira, insulin degludec and liraglutide, respectively 285.
FPG was similar with IDegLira (5.7 mmol/l) and insu-
lin degludec (6.0 mmol/l) by the end of the study, but
higher with liraglutide (7.3 mmol/l)285. The improve-
ments in glycaemic control were achieved with 37%
less daily insulin dose with IDegLira than with insulin
degludec285. IDegLira was associated with a significantly
greater decrease in body weight (estimated treatment
difference −2.80 kg, P <0.0001) and a 37% lower rate
of hypoglycaemia compared with insulin degludec285.
In patients who were already receiving basal insulin,
HbA1c decreased by 1.9% with IDegLira versus 0.9% with
insulin degludec (treatment difference −1.1%, 95% CI
−1.3% to −0.8%, P <0.0001). Mean weight reduction
with IDegLira was 2.7 kg versus no weight change with
insulin degludec, and the incidence of hypoglycaemia
was similar (24% for IDegLira versus 25% for insu-
lin degludec)286. Another fixed-ratio combination of
lixisenatide and insulin glargine has completed phase III
trials and a regulatory submission has been made
to the FDA287,288.
SGLT2 inhibitors
SGLT2 inhibitors currently available in Europe and North
America are dapagliflozin, canagliflozin and empag-
liflozin. They can be used as monotherapy when diet
and exercise are inadequate, and when metformin is not
tolerated, and can also be used as add-on to other glu-
cose-lowering agents, including insulin289. As their
efficacy is dependent on the renal filtration of glucose,
SGLT2 inhibitors should not be initiated in patients with
eGFR <60 ml/min/1.73 m2, but in patients who are
already receiving, and are tolerant of, canagliflozin or
empagliflozin, these medications can be continued in
patients with eGFR as low as 45 ml/min/1.73 m2 (REF. 290).
Mechanism of action
SGLTs are secondary active membrane symporters that
transfer sodium down its concentration gradient, usu-
ally into the cell, in conjunction with the transfer of
specific hexose sugars or other molecules against their
6CDNG| Head-to-head trials of glucagon-like peptide 1 receptor agonists (GLP-1RAs)166,177,179,,,194,195,,,s,,,,s
Study Design Baseline Background Active HbA1c change Weight Comments on
characteristics therapy comparators from baseline change from adverse effects
(%) baseline (kg)
• &74#6+10 • 41.#%0+ /GCPCIG|[GCTU &TWIPCKXGQT • 'ZGPCVKFG sXGTUWUŨ sXGTUWU *KIJGT
• 'ZGPCVKFG • n HbAEYGKIJV OGVHQTOKP OI39 P  ŨP KPEKFGPEGQH
39XGTUWU |YGGMU MI$/+MI 576<&QTC • 'ZGPCVKFG PCWUGCCPF
GZGPCVKFG$& O|[GCTU6&/ EQODKPCVKQP μI$& XQOKVKPIYKVJ
QHVYQQHVJQUG GZGPCVKFG
CIGPVU $&OQTG
KPLGEVKQPUKVG
TGCEVKQPUYKVJ
GZGPCVKFG39
• .'#& • 41.#%0+ /GCPCIG|[GCTU /GVHQTOKP57 • .KTCINWVKFG sXGTUWU sXGTUWU (GYGTCFXGTUG
• .KTCINWVKFG • n HbAEYGKIJV QTDQVJ OI3& ŨP ŨP GHHGEVUYKVJ
XGTUWU |YGGMU MI$/+MI • 'ZGPCVKFG NKTCINWVKFGDWV
GZGPCVKFG$& O|[GCTU6&/ μI$& QHVGPOQTG
UGXGTGVJCPYKVJ
GZGPCVKFG
• &74#6+10 • 41.#%0+ /GCPCIG|[GCTU &TWIPCKXGQT • 'ZGPCVKFG sXGTUWUŨ sXGTUWU 5KOKNCTVQ
• 'ZGPCVKFG • n HbAEYGKIJV OGVHQTOKP OI39 P  Ũ &74#6+10
39XGTUWU |YGGMU MI$/+MI 576<&QTCP[ • 'ZGPCVKFG PQPUKIPKHKECPV
GZGPCVKFG$& O|[GCTU6&/ EQODKPCVKQP μI$&
• &74#6+10 • 41.#%0+ /GCPCIG|[GCTU /GVHQTOKP • .KTCINWVKFG sXGTUWU sXGTUWU *KIJGTTCVGU
• .KTCINWVKFG • n HbAEYGKIJV 57DQVJQT OI3& ŨP Ũ QHPCWUGC
XGTUWU |YGGMU MI$/+MI OGVHQTOKP
 • 'ZGPCVKFG
RTGFGHKPGF P XQOKVKPICPF
GZGPCVKFG39 O|[GCTU6&/ RKQINKVC\QPG OI39 PQPKPHGTKQTKV[ FKCTTJQGCYKVJ
ETKVGTKCYGTGPQV NKTCINWVKFGOQTG
OGV KPLGEVKQPUKVG
TGCEVKQPUYKVJ
GZGPCVKFG39
• )GV)QCN: • 41.#%0+ /GCPCIG|[GCTU /GVHQTOKP • .KZKUGPCVKFG sXGTUWU sXGTUWU .QYGT
• .KZKUGPCVKFG • n HbAEYGKIJV μI3& ŨRTGFGHKPGF Ũ KPEKFGPEGUQH
XGTUWU |YGGMU MI$/+MI • 'ZGPCVKFG PQPKPHGTKQTKV[ PCWUGCCPF
GZGPCVKFG$& O|[GCTU6&/ μI$& ETKVGTKCYGTGOGV J[RQIN[ECGOKC
YKVJNKZKUGPCVKFG
VTGCVOGPV
• *#4/10; • 41.#%0+ /GCPCIG|[GCTU /GVHQTOKP • .KTCINWVKFG XGTUWU sXGTUWU 5NKIJVN[JKIJGT
• .KTCINWVKFG • n HbAEYGKIJV RKQINKVC\QPG OI3& P Ũ KPEKFGPEGQH
XGTUWU |YGGMU MI$/+MI 57QTCP[ • #NDKINWVKFG RTGFGHKPGF P  PCWUGCCPF
CNDKINWVKFG O|[GCTU6&/ EQODKPCVKQP OI39 PQPKPHGTKQTKV[ XQOKVKPIYKVJ
ETKVGTKCYGTGPQV NKTCINWVKFGOQTG
OGV KPLGEVKQPUKVG
TGCEVKQPUYKVJ
CNDKINWVKFG
• #9#4& • 41.2% /GCPCIG|[GCTU /GVHQTOKP
 • &WNCINWVKFG sXGTUWU sXGTUWU 0QFKHHGTGPEGU
• &WNCINWVKFG #%50+ HbAEYGKIJV RKQINKVC\QPG OI39 ŨXGTUWU
XGTUWU DGVYGGP
OIXGTUWU • n MI$/+MI • &WNCINWVKFG ŨXGTUWU ŨXGTUWU FWNCINWVKFGCPF
FWNCINWVKFG |YGGMU O|[GCTU6&/ OI39 ŨP 
 GZGPCVKFG
OIXGTUWU • 'ZGPCVKFG HQTDQVJ PHQT
GZGPCVKFG μI$& FWNCINWVKFG FWNCINWVKFG
$&XGTUWU • 2NCEGDQ FQUGUXGTUWU OIXGTUWU
RNCEGDQ GZGPCVKFG GZGPCVKFG
• #9#4& • 41.#%0+ /GCPCIG|[GCTU /GVHQTOKP • &WNCINWVKFG sXGTUWU sXGTUWU 0QFKHHGTGPEGU
• &WNCINWVKFG • n HbAEYGKIJV OI39 ŨRTGFGHKPGF Ũ DGVYGGPITQWRU
XGTUWU |YGGMU MI$/+MI • .KTCINWVKFG PQPKPHGTKQTKV[ P
NKTCINWVKFG O|[GCTU6&/ OI3& ETKVGTKCYGTGOGV
5WRGTKQTKV[VGUVKPIYCUXGTUWURNCEGDQPQPKPHGTKQTKV[VGUVKPIYCUXGTUWUGZGPCVKFG#%CEVKXGEQORCTCVQT$&VYKEGFCKN[0+PQPKPHGTKQTKV[1.QRGPNCDGN2%
RNCEGDQEQPVTQNNGF3&QPEGFCKN[39QPEGYGGMN[4TCPFQOK\GF5UWRGTKQTKV[57UWNHQP[NWTGC6&/V[RGFKCDGVGUOGNNKVWU6<&VJKC\QNKFKPGFKQPG

concentration gradient 291. SGLTs in the intestine and
kidneys transfer glucose across the luminal membrane
into enterocytes or ductal epithelial cells; glucose trans-
porters (GLUTs) mediate passive transfer of glucose
across basolateral membranes down its concentration
gradient289,292,293.
The main SGLTs are SGLT1 and SGLT2, which are
primarily responsible for intestinal glucose absorption
and for reabsorption of most of the filtered glucose in
the kidney, respectively 291,294. SGLT2 is a low-affinity,
high-capacity glucose transporter in the S1 segment of
the proximal tubules, which is suited to reabsorption
of a high concentration of filtered glucose entering the
tubules. SGLT1, which is also expressed in the kidneys,
is a high-affinity, low-capacity glucose transporter that is
suited to reabsorption of glucose at low concentration in
the S3 segment of the proximal tubules294–296.
Competitively inhibiting SGLT2 can eliminate
60–90 g glucose per day 297, but this amount can vary con-
siderably depending on renal function and the degree of
hyperglycaemia289. The effects of SGLT2 inhibition are
self-limiting, as the efficacy decreases as hyperglycaemia
lessens (and less glucose is filtered in the kidney). The
effects of SGLT2 inhibition are insulin-independent, and
efficacy is not affected by declining β-cell function or
insulin resistance28,289. However, insulin is still required,
as SGLT2 inhibition does not treat the underlying
endocrinopathies that contribute to the pathogenesis
of T2DM, except by reducing the effects of glucotoxic-
ity 28,289. SGLT2 inhibition and the associated glycosuria
result in mild diuresis and calorie loss, leading to mod-
est reductions in blood pressure and body weight 28,289.
However, the weight loss associated with SGLT2 inhibi-
tors is less than expected from the degree of glycosuria;
patients typically have one-quarter to one-third of the
weight loss predicted by their glycosuria. This effect is
partly accounted for by an elevation of calorie intake,
which correlates negatively with baseline BMI and pos-
itively with baseline eGFR298. In an RCT299 that included
95 patients who were taking a GLP-1RA (which should
counter increased calorie intake), addition of canagli-
flozin 300 mg resulted in significant weight loss com-
pared with placebo (LSMD for change in weight −3.2%,
95% CI −4.5% to −2.0%) over 18 weeks.
Pharmacokinetics
The currently available SGLT2 inhibitors have half-lives
of 10.6 h to 13.3 h289,300–304 (see Supplementary infor-
mation S7 (box)). Empagliflozin is the most specific.
SGLT2 inhibition by dapagliflozin (10 mg per day), can-
agliflozin (300 mg per day) or empagliflozin (25 mg per
day) increases urinary glucose excretion by 60–90 g per
day 289,302,303. SGLT2 inhibitors are metabolized by uridine
diphosphate glucuronosyl transferases, and no significant
interactions with other drugs have been reported289,305,306.
Pharmacodynamics
Dapagliflozin. In a 24-week RCT307 involving drug-naive
patients with T2DM, compared with placebo, dapagli-
flozin 5–10 mg per day reduced HbA1c by 0.8–0.9%, and
reduced body weight by 2.8–3.2 kg. A meta-analysis308
of RCTs of 12–104 weeks duration showed that dapagli-
flozin (2.5–10.0 mg per day) improved HbA1c, FPG and
weight compared with placebo when used as an add-on
therapy to metformin, insulin, thiazolidinediones, sul-
fonylureas or metformin ± sitagliptin (mean difference
between groups −0.52%, 95% CI −0.60% to −0.45%,
−1.52 mmol/l, 95% CI −1.75 mmol/l to −1.29 mmol/l and
−1.61 kg, 95% CI −1.97 kg to −1.26 kg, respectively). The
reductions in HbA1c and FPG compared with placebo
were generally similar with different background treat-
ments, but were greatest when dapagliflozin was added
to a sulfonylurea (−0.96%, 95% CI −0.86% to −0.52% and
−1.47 mmol/l, 95% CI −1.86 mmol/l to −1.08 mmol/l)308.
The largest between-group difference in weight change
was seen when dapagliflozin was added to insulin
(−2.45 kg, 95% CI −2.99 kg to −1.92 kg)308.
Dapagliflozin and glipizide were compared in a
52-week RCT with a 156-week extension309; dapagliflozin
resulted in lesser HbA1c reductions in the initial 18-week
titration phase, but the 18–104-week coefficient of fail-
ure was lower with dapagliflozin (0.13% per year) than
with glipizide (0.59% per year). HbA1c reductions were
greater with dapagliflozin by week 104 (difference from
glipizide −0.18%, 95% CI −0.33% to −0.03%, P = 0.021)309.
Dapagliflozin also resulted in sustained reductions in
weight and systolic blood pressure (104-week differences
from glipizide −5.1 kg, 95% CI −5.7 kg to −4.4 kg and
−3.9 mmHg, 95% CI −6.1 mmHg to −1.7 mmHg, respec-
tively)309. In an RCT310 involving 180 patients with T2DM
inadequately controlled by metformin, a modest level of
weight loss with dapagliflozin add-on compared with
placebo was associated with significant improvements in
health-related quality of life over 102 weeks. In an RCT311
involving 18 men with T2DM, in comparison with pla-
cebo dapagliflozin resulted in increased glucagon secre-
tion from as early as 1 h after administration, reaching a
peak after 4 h. After 3 days of dapagliflozin treatment, the
fasting plasma glucagon concentration was 32% higher
than on day 1, compared with no change in the placebo
group311. The increase in glucagon was associated with
increased endogenous glucose production311. The mech-
anism underlying this apparently compensatory change is
not known, although SGLT2 expression has been noted in
pancreatic α cells312.
Canagliflozin. In a meta-analysis of RCTs, canagliflozin
was found to reduce HbA1c when used as monotherapy
(WMD −1.08%, 95% CI −1.25% to −0.90%, P <0.00001)
or add-on treatment (−0.73%, 95% CI −0.84% to −0.61%,
P <0.00001), compared with placebo313. Relative to active
comparators, canagliflozin reduced HbA1c by −0.21%
(95% CI −0.33% to −0.08%, P = 0.001)313. Canagliflozin
also reduced HbA1c in comparison with sitagliptin
(−0.24%, 95% CI −0.40% to −0.09%, P = 0.002) and
glimepiride (−0.12%, 0.95% CI −0.23% to −0.01%,
P = 0.03)313, and reduced FPG in comparison with placebo
(−33.50 mg/dl, 95% CI −39.22 mg/dl to −27.78 mg/dl,
P <0.00001) and active comparators (−15.86 mg/dl,
95% CI −23.17 mg/dl to −8.56 mg/dl, P <0.00001)313.
Canagliflozin resulted in greater weight loss than pla-
cebo (−2.81 kg, 95% CI −3.26 kg to −2.37 kg) or active
comparators (−3.49 kg, 95% CI −4.86 kg to −2.12 kg),
particularly when compared with glimepiride (−5.40 kg,
95% CI −5.95 kg to −4.85 kg, P <0.00001)313.
Addition of canagliflozin to insulin treatment (gen-
erally by a basal–bolus regimen) has been shown to
result in a significant reduction in HbA1c at 18 weeks
(and sustained up to 52 weeks) compared with placebo,
from a baseline of 8.3% (−0.62%, 95% CI −0.69% to
−0.54%, P <0.001 with 100 mg canagliflozin and −0.73%,
95% CI −0.81% to −0.65%, P <0.001 with 300 mg can-
agliflozin)314. In an RCT involving 37 patients, follow-
ing an initial dose of canagliflozin, a second dose of
300 mg canagliflozin administered immediately before
a mixed-meal tolerance test reduced postprandial glu-
cose (compared with placebo) without causing further
increases in urinary glucose excretion, which suggests
the induction of mechanisms such as SGLT1 inhibition
in the gut 315. Glucose lowering and weight loss with can-
agliflozin are more durable than with sulfonylureas at
104 weeks316.
Reductions in systolic and diastolic blood pressure
have been demonstrated for canagliflozin compared with
placebo (systolic −5.05 mmHg, 95% CI −6.81 mmHg to
−3.28 mmHg, P <0.00001, diastolic −2.43 mmHg, 95% CI
−3.29 mmHg to −1.57 mmHg, P <0.0001) or active com-
parators (systolic −4.34 mmHg, 95% CI −5.31 mmHg
to −3.36 mmHg, P <0.00001, diastolic −2.17 mmHg,
95% CI −2.79 mmHg to −1.54 mmHg, P <0.00001)313.
Empagliflozin. In 24-week RCTs317–320, empagliflozin
(as monotherapy or added to metformin, to met-
formin with sulfonylurea or to pioglitazone with or
without metformin) resulted in reductions in HbA1c,
body weight and systolic blood pressure of 0.7–0.8%,
1.5–2.5 kg and 2.9–4.1 mmHg respectively, which were
significant in comparison with placebo. Reductions in
HbA1c and weight were maintained in trial extensions
up to 76 weeks321–324.
Compared with sitagliptin monotherapy for
24 weeks, empagliflozin monotherapy resulted in similar
HbA1c reductions, but greater reductions in FPG, body
weight and systolic blood pressure317. Over 104 weeks,
empagliflozin was noninferior to glimepiride as an
add-on to metformin treatment, and resulted in less
hypoglycaemia325.
In patients receiving basal insulin (with or without
the additional regimen of metformin with or without
sulfonylurea), empagliflozin resulted in HbA1c reduc-
tion of 2.0–2.5% compared with placebo over 78 weeks,
along with 2.4–4.1 kg weight loss326. Addition of pla-
cebo, empagliflozin 10 mg or empagliflozin 25 mg to a
multiple daily injection insulin regimen reduced HbA1c
(−0.81 ± 0.08%, −1.18 ± 0.08% and −1.27 ± 0.08%, respec-
tively) after 52 weeks327. Insulin dose and body weight
were reduced (by −9 to −11 international units per day
and −2.4 kg to −2.5 kg, respectively) with empagliflozin
treatment compared with placebo, without increasing
the risk of hypoglycaemia327.
In a 12-week RCT328 of patients with T2DM, base-
line systolic blood pressure of 130–159 mmHg and
diastolic blood pressure of 80–99 mmHg, the adjusted
mean difference versus placebo in change from baseline
in mean 24-h systolic blood pressure was −4.16 mmHg
(95% CI −5.50 mmHg to −2.83 mmHg), and in diastolic
blood pressure was −1.72 mmHg (95% CI −2.51 mmHg
to−0.93  mmHg) with 25 mg empagliflozin (both
P <0.001).
Compared with placebo, empagliflozin resulted in
an adjusted mean HbA1c difference of −0.68% (95% CI
−0.88% to −0.49%) in patients with eGFR 60–90 ml/
min/1.73 m2 and −0.42% (95% CI −0.56% to −0.28%) in
patients with eGFR 30–60 ml/min/1.73 m2 over 24 weeks,
and the treatment was well tolerated329.
Safety and adverse effects
SGLT2 inhibitors are associated with low risk of hypo-
glycaemia except when used in combination with insu-
lin or sulfonylureas289. This low risk reflects the ability
of uninhibited SGLT2 (and SGLT1) to reabsorb all of
a reduced filtered glucose load as the blood glucose
level declines, emphasizing the self-limiting nature of
this mode of action289. Compared with the sulfonylurea
glipizide, dapagliflozin resulted in significantly lower
risk of hypoglycaemia (4.2% versus 45.8%) in an RCT
involving 814 patients with T2DM over 104 weeks309.
Canagliflozin treatment was associated with similar rates
of hypoglycaemia to placebo when used as monotherapy
or as an add-on therapy, except when added to sulfo-
nylurea (RR 1.49, 95% CI 1.14–1.95, P = 0.004)313. The
percentage of patients having confirmed hypoglycaemic
events with empagliflozin treatment has been shown to
be <1% when used as monotherapy, 1.4–2.4% when used
as add-on to metformin or pioglitazone, 11.5–16.1%
when combined with sulfonylureas and 35–58% when
added to insulin28,239,305.
SGLT2 inhibitors are associated with increased risk of
genital infections, but an increase in urinary tract infection
(UTI) has not been consistently reported289. Compared
with sulfonylureas, dapagliflozin has been associated with
increased risk of genital and urinary tract infections
(14.8% and 13.5%, respectively, with dapagliflozin, 2.9%
and 9.1%, respectively, with glipizide)309. No increased
risk of UTIs was observed in patients treated with can-
agliflozin, but the risk of genital tract infections was
increased (RR 3.76, 95% CI 2.23–6.35, P <0.00001 versus
placebo, RR  4.95, 95%  CI 3.25–7.52, P <0.00001
versus active comparators); the increase was greater in
women than in men, but none of the reported infections
was severe and all were resolved with simple treatment313.
In a pooled analysis of RCTs330, genital mycotic infection
occurred more commonly with canagliflozin 100 mg and
300 mg than with placebo in women (10.4%, 11.4% and
3.2%, respectively) and in men (4.2%, 3.7% and 0.6%,
respectively). Similar results were found when canagli-
flozin was compared with active control (14.7%, 13.9%
and 3.1% in women, 7.3%, 9.3% and 1.6% in men)330.
The infections were generally mild and easy to treat, but
laboratory confirmation was lacking for most events330.
Similarly, a review of the properties of empagliflozin
found an association with UTI in some trials, but not
others, whereas all trials showed increased risk of genital
infections305.
SGLT2 inhibitors are associated with small increases
in LDL cholesterol, but also corresponding increases in
HDL cholesterol; these effects might be slightly greater
with canagliflozin than with other SGLT2 inhibitors313,331.
Results differ with regard to the risk of osmotic diuresis
and hypovolaemia289,332. In a meta-analysis313, the risks of
adverse effects related to osmotic diuresis were found to
be higher with canagliflozin than with placebo (RR 3.93,
95% CI 2.25–6.86, P <0.00001) or active comparators
(RR 2.57, 95% CI 1.26–5.25, P = 0.009), whereas vol-
ume-related adverse effects did not differ. In a 12-week
RCT333, canagliflozin 300 mg resulted in decreased plasma
volume at week 1 (−5.4% versus 4.3% with placebo,
P = 0.02), along with a modest increase in urinary volume,
both of which were attenuated by week 12. In a pooled
analysis334 of data from >11,000 patients with T2DM,
empagliflozin was not associated with the frequency of
events related to volume depletion, but a high frequency
of such events occurred in patients ≥75 years of age receiv-
ing empagliflozin 25 mg, and in patients receiving loop
diuretics in addition to empagliflozin 10 mg.
SGLT2 inhibitors, particularly canagliflozin, might
have adverse effects on the risk of fractures. The results
of an RCT335 with dapagliflozin showed no effect on
markers of bone formation or resorption, or bone min-
eral density after 50 weeks of treatment in men and post-
menopausal women with T2DM inadequately controlled
by metformin334,335. However, in some studies, canagli-
flozin has been shown to affect levels of urinary calcium,
serum phosphate and 1,25-dihydroxyvitamin D336. In a
26-week RCT336 with a 78-week extension that included
716 patients with T2DM aged 55–80 years, canagliflozin
treatment was associated with a decrease in total hip
bone mineral density over 104 weeks (placebo-subtracted
changes −0.9% and −1.2% for 100 mg and 300 mg canag-
liflozin, respectively), but no effect was seen at other bone
sites. In a pooled analysis of eight studies (n = 5,867), the
incidence of fractures was similar with (1.7%) and with-
out (1.5%) canagliflozin (HR 1.09, 95% CI 0.71–1.66)337.
Separate analysis of results from the CANVAS trial
(n = 4,327) showed a significant increase in fractures
with canagliflozin (4.0%) compared with placebo (2.6%;
HR 1.51, 95% CI 1.04–2.19), as well as increased fall-
related adverse effects337. However, compared with the
non-CANVAS trials, patients in the CANVAS trial were
older (62.4 ± 8.0 years versus 57.6 ± 9.8 years), with a high
risk of cardiovascular disease, and with lower baseline
eGFR and higher diuretic use337.
Several instances of euglycaemic and hyperglycaemic
diabetic ketoacidosis have been reported in patients who
received SGLT2 inhibitors338–341. The diabetic ketoacidosis
prevalence in 17,596 patients from randomized studies of
canagliflozin was 0.07% (n = 12)341. Many of the affected
patients, with T2DM treated with insulin, had reduced
or stopped insulin or experienced an intercurrent illness
that would increase the demand for glucose342. A lack
of insulin leads to increased lipolysis and conversion of
excess fatty acids to ketones, but the hyperglycaemia asso-
ciated with SGLT2 inhibitors is typically mild, presum-
ably because they reduce blood glucose338,339,342. In many
of the occurrences of diabetic ketoacidosis, reduction
of insulin dose revealed latent autoimmune diabetes of
adults, a form of type 1 diabetes mellitus (T1DM). Other
instances of diabetic ketoacidosis resulted from off-label
use of SGLT2 inhibitors in patients with T1DM338,339,342.
Patients treated with insulin and undertaking self-
monitoring of blood glucose should not, therefore, dis-
continue insulin when they observe a reduction in blood
glucose after introduction of an SGLT2 inhibitor. The
SGLT2 therapy can improve glycaemic control, but does
not obviate the need for insulin.
Pooled analysis of the results of phase II and phase III
trials suggests a beneficial effect of dapagliflozin on cardi-
ovascular disease65. Cardiovascular outcomes in patients
treated with SGLT2 inhibitors are being assessed in a
number of RCTs. In a study of 7,020 patients with T2DM
at high risk of cardiovascular events, occurrence of a
composite end point of nonfatal myocardial infarction,
nonfatal stroke and death from cardiovascular causes
was lower with empagliflozin than placebo, in addi-
tion to standard therapy (HR 0.86, 95% CI, 0.74–0.99,
P = 0.04 for superiority)343. Empagliflozin treatment
also reduced the risk of cardiovascular death (HR 0.62,
95% CI, 0.49–0.77, P <0.001), death from any cause
(HR 0.68, 95% CI, 0.57–0.82, P <0.001) and hospitali-
zation from heart failure (HR 0.65, 95% CI, 0.50–0.85,
P = 0.002)343. Subgroup analyses showed heterogeneity
for the primary outcome; the benefits of empagliflozin
were more evident in the Asian population, in patients
with BMI <30 kg/m2 and HbA1c <8.5%, in those not on
insulin treatment and in those with nephropathy 343. The
effect of empagliflozin on death from cardiovascular
causes was consistent across all subgroups343. Results of
other cardiovascular outcome trials with dapagliflozin
and canagliflozin are awaited with interest.
Other agents
Dopamine D2 receptor agonists
Bromocriptine quick release (QR) is a dopamine D2
receptor agonist that is licensed in some countries
outside Europe for treatment of T2DM as an adjunct
to lifestyle changes344,345. The effect of bromocriptine
on glycaemic parameters has been noted since 1980
(REF. 346). The drug provides a morning boost to hypo-
thalamic dopamine levels, consistent with normal diur-
nal glucoregulation, and contributing to a reduction of
sympathetic tone, neural suppression of hepatic glucose
production and improvement in peripheral glucose
disposal, without affecting insulin levels28,344,346,347. A
meta-analysis showed that bromocriptine QR add-on
therapy, compared with placebo, reduced levels of
HbA1c (−6.52  mmol/mol, 95%  CI −8.07 mmol/mol
to −4.97 mmol/mol) and FPG (−1.04 mmol/l, 95% CI
−1.49 mmol/l to −0.59 mmol/l), but had no effect on
postprandial glucose348. Bromocriptine QR was weight
neutral and was not associated with the risk of hypo-
glycaemia, hypotension or cardiovascular effects348.
However, bromocriptine QR increased gastrointesti-
nal adverse effects of nausea and vomiting, relative to
placebo348. In an RCT349 involving 3,095 patients, bro-
mocriptine QR (as monotherapy or add-on to glucose-
lowering agents, including insulin) was shown to reduce
the risk of cardiovascular disease, compared with placebo
(HR 0.60, 95% CI 0.35–0.96) by 52 weeks.
Bile-acid sequestrants
Bile-acid sequestrants are established treatments for
dyslipidaemia and are associated with a reduction in
the risk of cardiovascular disease350. In 2008, the FDA
licensed colesevelam as an adjunct to lifestyle measures
to improve glycaemic control in T2DM26. The mecha-
nism of action might involve the passage of bile acids
along the intestine, possibly activating bile-acid recep-
tors on L cells, leading to secretion of GLP-1. Inhibiting
the return of bile acids to the liver could also affect
glucose metabolism by preventing activation of hepatic
farnesoid receptors28. Colesevelam reduced HbA1c by
0.30–0.54% compared with placebo, in combination
with metformin, sulfonylureas, pioglitazone or insu-
lin, with no increased risk of hypoglycaemia or weight
gain350,351. Despite its favourable effect on levels of LDL
cholesterol and HDL cholesterol, colesevelam increased
levels of triglycerides by 11–22%350.
Pramlintide
Pramlintide, a soluble analogue of islet amyloid poly-
peptide, was introduced in 2005 as an injectable meal-
time adjunct to a basal–bolus insulin regimen352. It
assists glycaemic control and weight control through a
centrally-mediated effect via the area postrema, which
activates neural pathways that enhance satiety, suppress
pancreatic glucagon secretion and slow gastric empty-
ing 352. Modest reductions in HbA1c, typically 0.3–0.6%,
have been reported alongside body-weight reductions
of 1–2 kg and reductions of the bolus insulin require-
ment 352. Addition of pramlintide to treatment adds to
the burden of mealtime injections and requires care
with dose adjustments to minimize risks of nausea and
hypoglycaemia352.
Treatment algorithm
The treatment options for patients with T2DM now
extend to a variety of drug classes with different mecha-
nisms of action, low risks of hypoglycaemia and favour-
able effects on body weight. The availability of several
agents within most classes offers choice with regard to
pharmacokinetics, pharmacodynamics and the timing
and mode of delivery. However, direct comparisons can
be difficult when long-term head-to-head studies are not
available, as can determining suitability for individual
patients in the absence of studies in particular patient
subgroups. Overall, the choice of treatment must bal-
ance efficacy with safety, tolerability with adherence and
budgets with resources, as well as considering practical
issues relating to realistic targets, monitoring and life
situations36.
Metformin is firmly established as the preferred
first-line pharmacotherapy in patients with T2DM36.
Expectations are increasing for SGLT2 inhibitors, and
the results of ongoing RCTs will help to determine the
positioning of this class in the treatment algorithm.
Notably, the choice of metformin as first-line therapy
is mainly based on the results of the UKPDS, which
included 342 patients assigned to metformin, whereas
the efficacy of empagliflozin has been demonstrated in
4,687 patients in the EMPA–REG study 343. The EMPA–
REG study included patients with advanced disease and
high risk of cardiovascular disease, whereas the UKPDS
population had newly diagnosed T2DM. If HbA1c targets
are not met with metformin treatment within 3 months,
the recommendation from the American Diabetes
Association and the European Association for the
Study of Diabetes is to add a differently-acting agent 36.
Although oral agents will often have similar efficacy, the
injectables (GLP-1RAs and insulin) can have greater
effects on HbA1c241. However, efficacy is not just about
HbA1c, but must always take into account a ‘package’ of
effects that includes risk of hypoglycaemia, weight gain,
general tolerability and long-term safety. For example,
the risks of weight gain and hypoglycaemia are higher
with sulfonylureas and insulin than with DPP-4 inhibi-
tors and SGLT2 inhibitors36. Thiazolidinediones have a
low risk of hypoglycaemia, but increase body weight and
the risks of heart failure and bone fractures, compared
with placebo36. An individualised approach to treatment
is important, taking into account patients’ circumstances
and needs. Therapeutic choice is restricted in people
who drive, the elderly, the frail and those with renal,
neural and other comorbidities.
If the addition of a second agent fails to achieve or
maintain acceptable glycaemic control, adding a third
differently acting agent can be indicated36. Most classes of
agents can be combined with additive efficacy, although
addition of DPP-4 inhibitors to GLP-1RAs is unlikely
to offer extra control. If triple combinations are inad-
equate, introduction of insulin (usually basal initially
with continued metformin) is needed. If basal insulin is
insufficient, addition of meal-time insulin, a GLP-1RA
or possibly an SGLT2 inhibitor can be considered36.
Addition of a GLP-1RA in this context might be a useful
treatment strategy, as it has less risk of hypoglycaemia
than meal-time insulin, and has a better effect on weight.
The availability of increasing numbers of agents that
are given at a frequency less than daily might be attrac-
tive for many patients, and might enhance compliance.
The outcomes of ongoing cardiovascular safety studies
could further clarify the T2DM treatment algorithm,
as could the introduction of additional long-acting
GLP-1RAs, DPP-4 inhibitors and SGLT2 inhibitors that
are in development 18,26,353–357.
Lessons for future therapies
Advances in the understanding of the pathogenesis
of T2DM have informed the development of different
classes of treatments358. However, treatments are needed
with longer lasting metabolic effects than those currently
available, and with the ability to improve, or prevent con-
tinuing decline in, β-cell function. Clearly, safety is of
paramount importance. Adverse effects have been found
with several agents that have now been discontinued,
highlighting the importance of maintaining pharma-
covigilance. Minimizing hypoglycaemia, weight gain
and cardiovascular events while avoiding any increased
risk of cancer is crucial for new treatments, particularly
as they might need to be taken for many years. In real
life, medications will be used in more varied populations
than in clinical trials, and they might be prescribed by
less-specialized professionals to patients who will not
receive the intensive follow-up and monitoring associated
with RCTs359.
When considering safety, it can be extremely diffi-
cult to interpret results from preclinical studies, or to
have available the most appropriate models to decide
which treatments should be developed further. Another
challenge is to identify and interpret adverse signals in
clinical trials for extrapolation to real life359. Faint sig-
nals from preregistration trials can take a decade or
 more to reveal their clinical importance and are often
confounded by several biases, including treatment
allocation and detection of complications. Pressure to
ensure safety is increasing, but regulatory agencies have
a difficult task to strike a balance between appropri-
ate caution and making sure that new beneficial treat-
ments are made available in a safe, but timely manner 359.
Understanding the factors responsible for variations in
the responses of individuals to particular treatments,
and the influence of pharmacogenetics on pharma-
cokinetics and efficacy will facilitate personalized and
patient-centred therapies29.
Conclusion
Many different glucose-lowering therapies are now
available to address different aspects of the pathogenesis
of T2DM through a range of actions, and these treat-
ments vary in efficacy, convenience, adverse effect pro-
files and cost. The potential ‘value’ of a therapy involves
more than a cost–benefit analysis, and is based on a
‘package’ of attributes that takes account of long-term
safety, tolerability, risk of hypoglycaemia and weight
gain and suitability in the presence of comorbidities
and other medications. Individualized therapy must be
tailored to patients’ needs and preferences, with con-
sideration of their circumstances, understanding and
commitment.
DPP-4 inhibitors, GLP-1RAs and SGLT2 inhibitors
have low risks of hypoglycaemia (except when combined
with insulin or sulfonylurea) and are associated with
either weight loss or weight neutrality, but they are more
expensive than older agents such as sulfonylureas and
meglitinides. Evidence relating to the safety profiles of
many of these newer agents is encouraging and suggests
their value in the challenge to provide early, effective and
sustained glycaemic control in T2DM. Although met-
formin remains the preferred initial pharmacotherapy
(when tolerated), an individualized approach is required
to assess treatment targets and to achieve them in the
safest possible manner.

1. International Diabetes Federation. IDF Diabetes Atlas
7th edn http://www.diabetesatlas.org/ (2016).
2. de Groot, M., Anderson, R., Freedland, K. E.,
Clouse, R. E. & Lustman, P. J. Association of
depression and diabetes complications: a meta-
analysis. Psychosom. Med. 63, 619–630
(2001).
3. Jacobson, A. M. Impact of improved glycemic control
on quality of life in patients with diabetes. Endocr.
Pract. 10, 502–508 (2004).
4. Stumvoll, M., Goldstein, B. J. & van Haeften, T. W.
Type 2 diabetes: principles of pathogenesis and
therapy. Lancet 365, 1333–1346 (2005).
5. Kahn, S. E., Hull, R. L. & Utzschneider, K. M.
Mechanisms linking obesity to insulin resistance and
type 2 diabetes. Nature 444, 840–846
(2006).
6. Mulder, H., Nagorny, C. L., Lyssenko, V. & Groop, L.
Melatonin receptors in pancreatic islets: good
morning to a novel type 2 diabetes gene.
Diabetologia 52, 1240–1249 (2009).
7. Drucker, D. J. & Nauck, M. A. The incretin system:
glucagon-like peptide-1 receptor agonists and
dipeptidyl peptidase-4 inhibitors in type 2 diabetes.
Lancet 368, 1696–1705 (2006).
8. Wellen, K. E. & Hotamisligil, G. S. Inflammation,
stress, and diabetes. J. Clin. Invest. 115, 1111–1119
(2005).
9. Yang, Q. et al. Serum retinol binding protein 4
contributes to insulin resistance in obesity and
type 2 diabetes. Nature 436, 356–362
(2005).
10. Rui, L., Yuan, M., Frantz, D., Shoelson, S. &
White, M. F. SOCS-1 and SOCS-3 block insulin
signaling by ubiquitin-mediated degradation of IRS1
and IRS2. J. Biol. Chem. 277, 42394–42398
(2002).
11. Marcheva, B. et al. Disruption of the clock
components CLOCK and BMAL1 leads to
hypoinsulinaemia and diabetes. Nature 466,
627–631 (2010).
12. Yang, C. S. et al. Hypothalamic AMP-activated protein
kinase regulates glucose production. Diabetes 59,
2435–2443 (2010).
13. Marchetti, P. et al. in The Islets of Langerhans
(ed. Islam, M. S.) 501–514 (Springer Netherlands,
2010).
14. Campbell, J. E. & Drucker, D. J. Islet α cells and
glucagon — critical regulators of energy homeostasis.
Nat. Rev. Endocrinol. 11, 329–338 (2015).
15. Kahn, S. E. et al. Glycemic durability of rosiglitazone,
metformin, or glyburide monotherapy. N. Engl.
J. Med. 355, 2427–2443 (2006).
16. Defronzo, R. A. Banting Lecture. From the triumvirate
to the ominous octet: a new paradigm for the
treatment of type 2 diabetes mellitus. Diabetes 58,
773–795 (2009).
17. Bailey, C. J. Treating insulin resistance: future
prospects. Diab. Vasc. Dis. Res. 4, 20–31 (2007).
18. Altaf, Q. A., Barnett, A. H. & Tahrani, A. A. Novel
therapeutics for type 2 diabetes: insulin resistance.
Diabetes Obes. Metab. 17, 319–334 (2015).
19. Samuel, V. T. & Shulman, G. I. Mechanisms for insulin
resistance: common threads and missing links. Cell
148, 852–871 (2012).
20. Meier, B. C. & Wagner, B. K. Inhibition of HDAC3 as a
strategy for developing novel diabetes therapeutics.
Epigenomics 6, 209–214 (2014).
21. Brownlee, M. The pathobiology of diabetic
complications: a unifying mechanism. Diabetes 54,
1615–1625 (2005).
22. UK Prospective Diabetes Study (UKPDS) Group.
Intensive blood-glucose control with sulphonylureas
or insulin compared with conventional treatment and
risk of complications in patients with type 2 diabetes
(UKPDS 33). Lancet 352, 837–853 (1998).
23. Holman, R. R., Paul, S. K., Bethel, M. A.,
Matthews, D. R. & Neil, H. A. 10-year follow-up of
intensive glucose control in type 2 diabetes. N. Engl.
J. Med. 359, 1577–1589 (2008).
24. The Diabetes Control and Complications Trial
Research Group. The effect of intensive treatment of
diabetes on the development and progression of long-
term complications in insulin-dependent diabetes
mellitus. N. Engl. J. Med. 329, 977–986 (1993).
25. Hayward, R. A. et al. Follow-up of glycemic control and
cardiovascular outcomes in type 2 diabetes. N. Engl.
J. Med. 372, 2197–2206 (2015).
26. Tahrani, A. A., Bailey, C. J., Del Prato, S. &
Barnett, A. H. Management of type 2 diabetes: new
and future developments in treatment. Lancet 378,
182–197 (2011).
27. Tahrani, A. A., Piya, M. K., Kennedy, A. &
Barnett, A. H. Glycaemic control in type 2 diabetes:
targets and new therapies. Pharmacol. Ther. 125,
328–361 (2010).
28. Bailey, C. J. The current drug treatment landscape
for diabetes and perspectives for the future. Clin.
Pharmacol. Ther. 98, 170–184 (2015).
29. Maruthur, N. M. et al. The pharmacogenetics of
type 2 diabetes: a systematic review. Diabetes Care
37, 876–886 (2014).
30. Home, P. et al. Insulin therapy in people with type 2
diabetes: opportunities and challenges? Diabetes
Care 37, 1499–1508 (2014).
31. Ohkubo, Y. et al. Intensive insulin therapy prevents
the progression of diabetic microvascular
complications in Japanese patients with non-insulin-
dependent diabetes mellitus: a randomized
prospective 6-year study. Diabetes Res. Clin. Pract.
28, 103–117 (1995).
32. Dormandy, J. A. et al. Secondary prevention of
macrovascular events in patients with type 2
diabetes in the PROactive Study (PROspective
pioglitAzone Clinical Trial In macroVascular Events):
a randomised controlled trial. Lancet 366,
1279–1289 (2005).
33. Action to Control Cardiovascular Risk in
Diabetes Study Group. Effects of intensive glucose
lowering in type 2 diabetes. N. Engl. J. Med. 358,
2545–2559 (2008).
34. ADVANCE Collaborative Group. Intensive blood
glucose control and vascular outcomes in patients with
type 2 diabetes. N. Engl. J. Med. 358, 2560–2572
(2008).
35. Duckworth, W. et al. Glucose control and vascular
complications in veterans with type 2 diabetes.
N. Engl. J. Med. 360, 129–139 (2009).
36. Inzucchi, S. E. et al. Management of hyperglycemia in
type 2 diabetes, 2015: a patient-centered approach:
update to a position statement of the American
Diabetes Association and the European Association
for the Study of Diabetes. Diabetes Care 38,
140–149 (2015).
37. Krentz, A. J. & Bailey, C. J. Oral antidiabetic agents:
current role in type 2 diabetes mellitus. Drugs 65,
385–411 (2005).
38. Tahrani, A. A., Varughese, G. I., Scarpello, J. H. &
Hanna, F. W. F. Metformin, heart failure, and lactic
acidosis: is metformin absolutely contraindicated?
BMJ 335, 508–512 (2007).
39. Bailey, C. J. & Turner, R. C. Metformin. N. Engl.
J. Med. 334, 574–579 (1996).
40. Ferrannini, E. The target of metformin in type 2
diabetes. N. Engl. J. Med. 371, 1547–1548
(2014).
41. Zhou, G. et al. Role of AMP-activated protein kinase in
mechanism of metformin action. J. Clin. Invest. 108,
1167–1174 (2001).
42. Viollet, B. et al. Cellular and molecular mechanisms of
metformin: an overview. Clin. Sci. (Lond.) 122,
253–270 (2012).
43. Bailey, C. J., Wilcock, C. & Scarpello, J. H. Metformin
and the intestine. Diabetologia 51, 1552–1553
(2008).
44. Madiraju, A. K. et al. Metformin suppresses
gluconeogenesis by inhibiting mitochondrial
glycerophosphate dehydrogenase. Nature 510,
542–546 (2014).
45. Buse, J. B. et al. The primary glucose-lowering effect of
metformin resides in the gut, not the circulation:
results from short-term pharmacokinetic and 12-week
dose-ranging studies. Diabetes Care 39, 198–205
(2016).
46. Lindsay, J. R. et al. Inhibition of dipeptidyl peptidase
IV activity by oral metformin in type 2 diabetes.
Diabet. Med. 22, 654–657 (2005).
47. Mannucci, E. et al. Effect of metformin on glucagon-
like peptide 1 (GLP-1) and leptin levels in obese
nondiabetic subjects. Diabetes Care 24, 489–494
(2001).
48. Mannucci, E. et al. Effects of metformin on glucagon-
like peptide-1 levels in obese patients with and
without type 2 diabetes. Diabetes Nutr. Metab. 17,
336–342 (2004).
49. Maida, A., Lamont, B. J., Cao, X. & Drucker, D. J.
Metformin regulates the incretin receptor axis via a
pathway dependent on peroxisome proliferator-
activated receptor-α in mice. Diabetologia 54,
339–349 (2011).
50. Cho, Y. M. & Kieffer, T. J. New aspects of an old drug:
metformin as a glucagon-like peptide 1 (GLP-1)
enhancer and sensitiser. Diabetologia 54, 219–222
(2011).
51. Mulherin, A. J. et al. Mechanisms underlying
metformin-induced secretion of glucagon-like
peptide-1 from the intestinal L cell. Endocrinology
152, 4610–4619 (2011).
52. Um, J. H. et al. Activation of 5ʹ-AMP-activated
kinase with diabetes drug metformin induces casein
kinase Iε (CKIε)-dependent degradation of clock
protein mPer2. J. Biol. Chem. 282, 20794–20798
(2007).
53. Barnea, M. et al. Metformin affects the circadian
clock and metabolic rhythms in a tissue-specific
manner. Biochim. Biophys. Acta 1822, 1796–1806
(2012).
54. Scheen, A. J. Clinical pharmacokinetics of metformin.
Clin. Pharmacokinet. 30, 359–371 (1996).
55. Grant, P. J. Beneficial effects of metformin on
haemostasis and vascular function in man. Diabetes
Metab. 29, 6S44–6S52 (2003).
56. Del Prato, S., Bianchi, C. & Marchetti, P. β-cell function
and anti-diabetic pharmacotherapy. Diabetes Metab.
Res. Rev. 23, 518–527 (2007).
57. Dujic, T. et al. Association of organic cation
transporter 1 with intolerance to metformin in type 2
diabetes: a GoDARTS study. Diabetes 64,
1786–1793 (2015).
58. Heaf, J. Metformin in chronic kidney disease: time for
a rethink. Perit. Dial. Int. 34, 353–357 (2014).
59. Scheen, A. J. & Paquot, N. Metformin revisited: a
critical review of the benefit–risk balance in at-risk
patients with type 2 diabetes. Diabetes Metab. 39,
179–190 (2013).
60. Eurich, D. T. et al. Comparative safety and
effectiveness of metformin in patients with diabetes
mellitus and heart failure: systematic review of
observational studies involving 34,000 patients.
Circ. Heart Fail. 6, 395–402 (2013).
61. Hitchings, A. W., Archer, J. R. Srivastava, S. A. &
Baker, E. H. Safety of metformin in patients with
chronic obstructive pulmonary disease and type 2
diabetes mellitus. COPD 12, 126–131 (2015).
62. Hung, S. C. et al. Metformin use and mortality in
patients with advanced chronic kidney disease:
national, retrospective, observational, cohort study.
Lancet Diabetes Endocrinol. 3, 605–614 (2015).
63. Salpeter, S. R., Greyber, E., Pasternak, G. A. &
Salpeter, E. E. Risk of fatal and nonfatal lactic
acidosis with metformin use in type 2 diabetes
mellitus. Cochrane Database Syst. Rev. 4,
CD002967 (2010).
64. UK Prospective Diabetes Study (UKPDS) Group.
Effect of intensive blood-glucose control with
metformin on complications in overweight patients
with type 2 diabetes (UKPDS 34). Lancet 352,
854–865 (1998).
65. Ferrannini, E. & DeFronzo, R. A. Impact of glucose-
lowering drugs on cardiovascular disease in type 2
diabetes. Eur. Heart J. 36, 2288–2296 (2015).
66. Henquin, J. C. The fiftieth anniversary of
hypoglycaemic sulphonamides. How did the mother
compound work? Diabetologia 35, 907–912 (1992).
67. Ashcroft, F. M. & Gribble, F. M. ATP-sensitive K+
channels and insulin secretion: their role in health and
disease. Diabetologia 42, 903–919 (1999).
68. Groop, L. C. Sulfonylureas in NIDDM. Diabetes Care
15, 737–754 (1992).
69. Lebovitz, H. E. Insulin secretagogues: old and new.
Diabetes Rev. 7, 139–153 (1999).
70. Rendell, M. The role of sulphonylureas in the
management of type 2 diabetes mellitus. Drugs 64,
1339–1358 (2004).
71. Abe, M., Okada, K. & Soma, M. Antidiabetic agents in
patients with chronic kidney disease and end-stage
renal disease on dialysis: metabolism and clinical
practice. Curr. Drug Metab. 12, 57–69 (2011).
72. Scheen, A. J. Pharmacokinetic considerations for the
treatment of diabetes in patients with chronic kidney
disease. Expert Opin. Drug Metab. Toxicol. 9,
529–550 (2013).
73. Arnouts, P. et al. Glucose-lowering drugs in patients
with chronic kidney disease: a narrative review on
pharmacokinetic properties. Nephrol. Dial. Transplant.
29, 1284–1300 (2014).
74. Schernthaner, G. et al. GUIDE study: double-blind
comparison of once-daily gliclazide MR and
glimepiride in type 2 diabetic patients. Eur. J. Clin.
Invest. 34, 535–542 (2004).
75. Carlson, R. F., Isley, W. L., Ogrinc, F. G. &
Klobucar, T. R. Efficacy and safety of reformulated,
micronized glyburide tablets in patients with non-
insulin-dependent diabetes mellitus: a multicenter,
double-blind, randomized trial. Clin. Ther. 15,
788–796 (1993).
76. Chung, M. et al. Pharmacokinetics and
pharmacodynamics of extended-release glipizide GITS
compared with immediate-release glipizide in patients
with type II diabetes mellitus. J. Clin. Pharmacol. 42,
651–657 (2002).
77. UK Hypoglycaemia Study Group. Risk of
hypoglycaemia in types 1 and 2 diabetes: effects of
treatment modalities and their duration. Diabetologia
50, 1140–1147 (2007).
78. Barnett, A. H. et al. The efficacy of self-monitoring of
blood glucose in the management of patients with
type 2 diabetes treated with a gliclazide modified
release-based regimen. A multicentre, randomized,
parallel-group, 6-month evaluation (DINAMIC 1
study). Diabetes Obes. Metab. 10, 1239–1247
(2008).
79. Meinert, C. L., Knatterud, G. L., Prout, T. E. &
Klimt, C. R. A study of the effects of hypoglycemic
agents on vascular complications in patients with
adult-onset diabetes. II. Mortality results. Diabetes 19
(Suppl.), 789–830 (1970).
80. US National Library of Science. ClinicalTrials.gov
https://clinicaltrials.gov/ct2/show/NCT01243424
(2016).
81. Campbell, I. W. Nateglinide — current and future role
in the treatment of patients with type 2 diabetes
mellitus. Int. J. Clin. Pract. 59, 1218–1228 (2005).
82. Dornhorst, A. Insulinotropic meglitinide analogues.
Lancet 358, 1709–1716 (2001).
83. Guardado-Mendoza, R., Prioletta, A., Jiménez-
Ceja, L. M., Sosale, A. & Folli, F. The role of nateglinide
and repaglinide, derivatives of meglitinide, in the
treatment of type 2 diabetes mellitus. Arch. Med. Sci.
9, 936–943 (2013).
84. Scheen, A. J. Drug–drug and food–drug
pharmacokinetic interactions with new insulinotropic
agents repaglinide and nateglinide. Clin.
Pharmacokinet. 46, 93–108 (2007).
85. Panelo, A. & Wing, J. R. Repaglinide/bedtime NPH
insulin is comparable to twice-daily NPH insulin.
Diabetes Care 28, 1789–1790 (2005).
86. Lund, S. S. et al. Combining insulin with metformin or
an insulin secretagogue in non-obese patients with
type 2 diabetes: 12 month, randomised, double blind
trial. BMJ 339, b4324 (2009).
87. Rosenstock, J. et al. Repaglinide versus nateglinide
monotherapy: a randomized, multicenter study.
Diabetes Care 27, 1265–1270 (2004).
88. Madsbad, S., Kilhovd, B., Lager, I., Mustajoki, P. &
Dejgaard, A. Comparison between repaglinide and
glipizide in type 2 diabetes mellitus: a 1-year
multicentre study. Diabet. Med. 18, 395–401
(2001).
89. Marbury, T., Huang, W. C., Strange, P. & Lebovitz, H.
Repaglinide versus glyburide: a one-year comparison
trial. Diabetes Res. Clin. Pract. 43, 155–166 (1999).
90. Meneilly, G. S. Effect of repaglinide versus glyburide
on postprandial glucose and insulin values in elderly
patients with type 2 diabetes. Diabetes Technol. Ther.
13, 63–65 (2011).
91. Schwarz, S. L. et al. Nateglinide, alone or in
combination with metformin, is effective and well
tolerated in treatment naïve elderly patients with
type 2 diabetes. Diabetes Obes. Metab. 10, 652–660
(2008).
92. Bellomo Damato, A., Stefanelli, G., Laviola, L.,
Giorgino, R. & Giorgino, F. Nateglinide provides tighter
glycaemic control than glyburide in patients with
type 2 diabetes with prevalent postprandial
hyperglycaemia. Diabet. Med. 28, 560–566 (2011).
93. Hu, S. et al. Pancreatic ß-cell K(ATP) channel activity
and membrane-binding studies with nateglinide: a
comparison with sulfonylureas and repaglinide.
J. Pharmacol. Exp. Ther. 293, 444–452 (2000).
94. NAVIGATOR Study Group et al. Effect of nateglinide on
the incidence of diabetes and cardiovascular events.
N. Engl. J. Med. 362, 1463–1476 (2010).
95. Huang, Y., Abdelmoneim, A. S., Light, P., Qiu, W. &
Simpson, S. H. Comparative cardiovascular safety of
insulin secretagogues following hospitalization for
ischemic heart disease among type 2 diabetes patients:
a cohort study. J. Diabetes Compl. 29, 196–202 (2015).
96. Bailey, C. J. Diabetes Therapies: Treating
Hyperglycaemia 77–82 (MedEd UK Limited, 2009).
97. Ueno, H. et al. Effects of miglitol, acarbose, and
sitagliptin on plasma insulin and gut peptides in
type 2 diabetes mellitus: a crossover study. Diabetes
Ther. 6, 187–196 (2015).
98. Joshi, S. R. et al. Therapeutic potential of
α-glucosidase inhibitors in type 2 diabetes mellitus: an
evidence-based review. Expert Opin. Pharmacother.
16, 1959–1981 (2015).
99. Yang, W. et al. Acarbose compared with metformin as
initial therapy in patients with newly diagnosed type 2
diabetes: an open-label, non-inferiority randomised
trial. Lancet Diabetes Endocrinol. 2, 46–55 (2014).
100. van de Laar, F. A. et al. Is acarbose equivalent to
tolbutamide as first treatment for newly diagnosed
type 2 diabetes in general practice? A randomised
controlled trial. Diabetes Res. Clin. Pract. 63, 57–65
(2004).
101. Chiasson, J. L. et al. Acarbose for prevention of type 2
diabetes mellitus: the STOP-NIDDM randomised trial.
Lancet 359, 2072–2077 (2002).
102. Chiasson, J. L. et al. Acarbose treatment and the risk
of cardiovascular disease and hypertension in patients
with impaired glucose tolerance: the STOP-NIDDM
trial. JAMA 290, 486–494 (2003).
103. Holman, R. R. et al. Rationale for and design of the
Acarbose Cardiovascular Evaluation (ACE) trial. Am.
Heart J. 168, 23–29 (2014).
104. Yki-Järvinen, H. Thiazolidinediones. N. Engl. J. Med.
351, 1106–1118 (2004).
105. Hauner, H. The mode of action of thiazolidinediones.
Diabetes Metab. Res. Rev. 18, S10–S15 (2002).
106. Scheen, A. J. Pharmacokinetic interactions with
thiazolidinediones. Clin. Pharmacokinet. 46, 1–12
(2007).
107. Rosenstock, J., Einhorn, D., Hershon, K., Glazer, N. B.
& Yu, S. Efficacy and safety of pioglitazone in type 2
diabetes: a randomised, placebo-controlled study in
patients receiving stable insulin therapy. Int. J. Clin.
Pract. 56, 251–257 (2002).
108. Charbonnel, B. et al. Long-term efficacy and
tolerability of add-on pioglitazone therapy to failing
monotherapy compared with addition of gliclazide or
metformin in patients with type 2 diabetes.
Diabetologia 48, 1093–1104 (2005).
109. Punthakee, Z. et al. Impact of rosiglitazone on body
composition, hepatic fat, fatty acids, adipokines and
glucose in persons with impaired fasting glucose or
impaired glucose tolerance: a sub-study of the DREAM
trial. Diabet. Med. 31, 1086–1092 (2014).
110. Loke, Y. K., Singh, S. & Furberg, C. D. Long-term use
of thiazolidinediones and fractures in type 2 diabetes:
a meta-analysis. CMAJ 180, 32–39 (2009).
111. Schwartz, A. V. et al. Effects of TZD use and
discontinuation on fracture rates in ACCORD bone study.
J. Clin. Endocrinol. Metab. 100, 4059–4066 (2015).
112. Billington, E. O., Grey, A. & Bolland, M. J. The effect of
thiazolidinediones on bone mineral density and bone
turnover: systematic review and meta-analysis.
Diabetologia 58, 2238–2246 (2015).
113. Nissen, S. E. & Wolski, K. Effect of rosiglitazone on the
risk of myocardial infarction and death from
cardiovascular causes. N. Engl. J. Med. 356,
2457–2471 (2007).
114. Mahaffey, K. W. et al. Results of a reevaluation of
cardiovascular outcomes in the RECORD trial.
Am. Heart J. 166, 240–249.e1 (2013).
115. Winkler, K. et al. Pioglitazone reduces atherogenic
dense LDL particles in nondiabetic patients with arterial
hypertension. Diabetes Care 26, 2588–2594 (2003).
116. Erdmann, E. et al. The effect of pioglitazone on
recurrent myocardial infarction in 2,445 patients with
type 2 diabetes and previous myocardial infarction:
results from the PROactive (PROactive 05) Study.
J. Am. Coll. Cardiol. 49, 1772–1780 (2007).
117. Wilcox, R. et al. Effects of pioglitazone in patients with
type 2 diabetes with or without previous stroke results
from PROactive (PROspective pioglitAzone Clinical
Trial In macroVascular Events 04). Stroke 38,
865–873 (2007).
118. Pfister, R., Cairns, R., Erdmann, E. & Schneider, C. A.
A clinical risk score for heart failure in patients with
type 2 diabetes and macrovascular disease: an
analysis of the PROactive study. Int. J. Cardiol. 162,
112–116 (2013).
119. Nissen, S. E. et al. Comparison of pioglitazone versus
glimepiride on progression of coronary atherosclerosis
in patients with type 2 diabetes: the PERISCOPE
randomized controlled trial. JAMA 299, 1561–1573
(2008).
120. Mazzone, T. et al. Effect of pioglitazone compared with
glimepiride on carotid intima-media thickness in
type 2 diabetes: a randomized trial. JAMA 296,
2572–2581 (2006).
121. Hughes, A. D. et al. A randomized placebo controlled
double blind crossover study of pioglitazone on left
ventricular diastolic function in type 2 diabetes. Int.
J. Cardiol. 167, 1329–1332 (2013).
122. Straznicky, N. E. et al. A randomized controlled trial of
the effects of pioglitazone treatment on sympathetic
nervous system activity and cardiovascular function in
obese subjects with metabolic syndrome. J. Clin.
Endocrinol. Metab. 99, E1701–E1707 (2014).
123. Palalau, A. I., Tahrani, A. A., Piya, M. K. &
Barnett, A. H. DPP-4 inhibitors in clinical practice.
Postgrad. Med. 121, 70–100 (2009).
124. McKeage, K. Trelagliptin: first global approval. Drugs
75, 1161–1164 (2015).
125. Burness, C. B. Omarigliptin: first global approval.
Drugs 75, 1947–1952 (2015).
126. Elrick, H., Stimmler, L., Hlad, C. J. & Arai, Y. Plasma
insulin response to oral and intravenous glucose
administration. J. Clin. Endocrinol. Metab. 24,
1076–1082 (1964).
127. Nauck, M. A. et al. Incretin effects of increasing
glucose loads in man calculated from venous insulin
and C-peptide responses. J. Clin. Endocrinol. Metab.
63, 492–498 (1986).
128. Baggio, L. L. & Drucker, D. J. Biology of incretins:
GLP-1 and GIP. Gastroenterology 132, 2131–2157
(2007).
129. Fehmann, H. C., Göke, R. & Göke, B. Cell and
molecular biology of the incretin hormones glucagon-
like peptide-I and glucose-dependent insulin
releasing polypeptide. Endocr. Rev. 16, 390–410
(1995).
130. Gautier, J. F., Choukem, S. P. & Girard, J. Physiology of
incretins (GIP and GLP-1) and abnormalities in type 2
diabetes. Diabetes Metab. 34, S65–S72 (2008).
131. Dupre, J., Ross, S. A., Watson, D. & Brown, J. C.
Stimulation of insulin secretion by gastric inhibitory
polypeptide in man. J. Clin. Endocrinol. Metab. 37,
826–828 (1973).
132. Trümper, A. et al. Glucose-dependent insulinotropic
polypeptide is a growth factor for β (INS-1) cells by
pleiotropic signaling. Mol. Endocrinol. 15,
1559–1570 (2001).
133. Yip, R. G. & Wolfe, M. M. GIP biology and fat
metabolism. Life Sci. 66, 91–103 (2000).
134. Ahrén, B. Gut peptides and type 2 diabetes mellitus
treatment. Curr. Diab. Rep. 3, 365–372 (2003).
135. Barnett, A. H. New treatments in type 2 diabetes: a
focus on the incretin-based therapies. Clin. Endocrinol.
(Oxf.) 70, 343–353 (2009).
136. Drucker, D. J., Philippe, J., Mojsov, S., Chick, W. L. &
Habener, J. F. Glucagon-like peptide I stimulates
insulin gene expression and increases cyclic AMP
levels in a rat islet cell line. Proc. Natl Acad. Sci. USA
84, 3434–3438 (1987).
137. Drucker, D. J. Glucagon-like peptides: regulators of
cell proliferation, differentiation, and apoptosis. Mol.
Endocrinol. 17, 161–171 (2003).
138. Malmgren, S. & Ahrén, B. DPP-4 inhibition
contributes to the prevention of hypoglycaemia
through a GIP–glucagon counterregulatory axis in
mice. Diabetologia 58, 1091–1099 (2015).
139. Farngren, J., Persson, M., Schweizer, A., Foley, J. E. &
Ahrén, B. Vildagliptin reduces glucagon during
hyperglycemia and sustains glucagon
counterregulation during hypoglycemia in type 1
diabetes. J. Clin. Endocrinol. Metab. 97, 3799–3806
(2012).
140. Hansen, L., Deacon, C. F., Orskov, C. & Holst, J. J.
Glucagon-like peptide-1-(7–36)amide is transformed
to glucagon-like peptide-1-(9–36)amide by dipeptidyl
peptidase IV in the capillaries supplying the L cells of
the porcine intestine. Endocrinology 140,
5356–5363 (1999).
141. Deacon, C. F., Johnsen, A. H. & Holst, J. J.
Degradation of glucagon-like peptide-1 by human
plasma in vitro yields an N-terminally truncated
peptide that is a major endogenous metabolite in vivo.
J. Clin. Endocrinol. Metab. 80, 952–957 (1995).
142. Mentlein, R. Dipeptidyl-peptidase IV (CD26) — role in
the inactivation of regulatory peptides. Regul. Pept.
85, 9–24 (1999).
143. He, Y. L. et al. Pharmacokinetics and
pharmacodynamics of vildagliptin in patients with
type 2 diabetes mellitus. Clin. Pharmacokinet. 46,
577–588 (2007).
144. Herman, G. A. et al. Pharmacokinetics and
pharmacodynamics of sitagliptin, an inhibitor of
dipeptidyl peptidase IV, in healthy subjects: results
from two randomized, double-blind, placebo-
controlled studies with single oral doses. Clin.
Pharmacol. Ther. 78, 675–688 (2005).
145. Nauck, M. A. & El-Ouaghlidi, A. The therapeutic
actions of DPP-IV inhibition are not mediated by
glucagon-like peptide-1. Diabetologia 48, 608–611
(2005).
146. Mearns, E. S. et al. Comparative efficacy and safety of
antidiabetic drug regimens added to metformin
monotherapy in patients with type 2 diabetes: a
network meta-analysis. PLoS ONE 10, e0125879
(2015).
147. Day, C. & Bailey, C. J. Pharmacotherapies to manage
diabesity: an update. Diabes. Pract. 4, 14–23 (2015).
148. Tahrani, A. A., Piya, M. K. & Barnett, A. H.
Saxagliptin: a new DPP-4 inhibitor for the treatment
of type 2 diabetes mellitus. Adv. Ther. 26, 249–262
(2009).
149. Tahrani, A. A., Piya, M. K. & Barnett, A. H. Drug
evaluation: vildagliptin-metformin single-tablet
combination. Adv. Ther. 26, 138–154 (2009).
150. Deacon, C. F. & Lebovitz, H. E. Comparative review of
dipeptidyl peptidase-4 inhibitors and sulphonylureas.
Diabetes Obes. Metab. 18, 333–347 (2016).
151. Deacon, C. F. Dipeptidyl peptidase-4 inhibitors in the
treatment of type 2 diabetes: a comparative review.
Diabetes Obes. Metab. 13, 7–18 (2011).
152. Scheen, A. J. Dipeptidylpeptidase-4 inhibitors
(gliptins). Clin. Pharmacokinet. 49, 573–588 (2010).
153. Esposito, K. et al. A nomogram to estimate the HbA1c
response to different DPP-4 inhibitors in type 2
diabetes: a systematic review and meta-analysis of 98
trials with 24 163 patients. BMJ Open 5, e005892
(2015).
154. Karagiannis, T., Paschos, P., Paletas, K.,
Matthews, D. R. & Tsapas, A. Dipeptidyl peptidase-4
inhibitors for treatment of type 2 diabetes mellitus in
the clinical setting: systematic review and meta-
analysis. BMJ 344, e1369 (2012).
155. Zhang, Y. et al. Head-to-head comparison of dipeptidyl
peptidase-IV inhibitors and sulfonylureas — a meta-
analysis from randomized clinical trials. Diabetes
Metab. Res. Rev. 30, 241–256 (2014).
156. Derosa, G. et al. A randomized, double-blind,
comparative therapy evaluating sitagliptin versus
glibenclamide in type 2 diabetes patients already
treated with pioglitazone and metformin: a 3-year
study. Diabetes Technol. Ther. 15, 214–222 (2013).
157. Rosenstock, J., Wilson, C. & Fleck, P. Alogliptin versus
glipizide monotherapy in elderly type 2 diabetes
mellitus patients with mild hyperglycaemia: a
prospective, double-blind, randomized, 1-year study.
Diabetes Obes. Metab. 15, 906–914 (2013).
158. Gallwitz, B. et al. 2-year efficacy and safety of
linagliptin compared with glimepiride in patients with
type 2 diabetes inadequately controlled on metformin:
a randomised, double-blind, non-inferiority trial.
Lancet 380, 475–483 (2012).
159. Filozof, C. & Gautier, J. F. A comparison of efficacy and
safety of vildagliptin and gliclazide in combination with
metformin in patients with type 2 diabetes
inadequately controlled with metformin alone: a
52-week, randomized study. Diabet. Med. 27,
318–326 (2010).
160. Matthews, D. R. et al. Vildagliptin add-on to metformin
produces similar efficacy and reduced hypoglycaemic
risk compared with glimepiride, with no weight gain:
results from a 2-year study. Diabetes Obes. Metab.
12, 780–789 (2010).
161. Foley, J. E. & Sreenan, S. Efficacy and safety
comparison between the DPP-4 inhibitor vildagliptin
and the sulfonylurea gliclazide after two years of
monotherapy in drug-naive patients with type 2
diabetes. Horm. Metab. Res. 41, 905–909 (2009).
162. Arjona Ferreira, J. C. et al. Efficacy and safety of
sitagliptin in patients with type 2 diabetes and ESRD
receiving dialysis: a 54-week randomized trial. Am.
J. Kidney Dis. 61, 579–587 (2013).
163. Arjona Ferreira, J. C. et al. Efficacy and safety of
sitagliptin versus glipizide in patients with type 2
diabetes and moderate-to-severe chronic renal
insufficiency. Diabetes Care 36, 1067–1073 (2013).
164. Seck, T. et al. Safety and efficacy of treatment with
sitagliptin or glipizide in patients with type 2 diabetes
inadequately controlled on metformin: a 2-year study.
Int. J. Clin. Pract. 64, 562–576 (2010).
165. Kim, Y. G. et al. Differences in the glucose-lowering
efficacy of dipeptidyl peptidase-4 inhibitors between
Asians and non-Asians: a systematic review and meta-
analysis. Diabetologia 56, 696–708 (2013).
166. Zimdahl, H. et al. Influence of TCF7L2 gene variants
on the therapeutic response to the
dipeptidylpeptidase-4 inhibitor linagliptin.
Diabetologia 57, 1869–1875 (2014).
167. Esposito, K. et al. Glycaemic durability with dipeptidyl
peptidase-4 inhibitors in type 2 diabetes: a systematic
review and meta-analysis of long-term randomised
controlled trials. BMJ Open 4, e005442 (2014).
168. Scheen, A. J., Charpentier, G., Ostgren, C. J.,
Hellqvist, A. & Gause-Nilsson, I. Efficacy and safety of
saxagliptin in combination with metformin compared
with sitagliptin in combination with metformin in adult
patients with type 2 diabetes mellitus. Diabetes
Metab. Res. Rev. 26, 540–549 (2010).
169. Kothny, W., Lukashevich, V., Foley, J. E., Rendell, M. S.
& Schweizer, A. Comparison of vildagliptin and
sitagliptin in patients with type 2 diabetes and severe
renal impairment: a randomised clinical trial.
Diabetologia 58, 2020–2026 (2015).
170. Inagaki, N., Onouchi, H., Maezawa, H., Kuroda, S. &
Kaku, K. Once-weekly trelagliptin versus daily alogliptin
in Japanese patients with type 2 diabetes: a
randomised, double-blind, phase 3, non-inferiority
study. Lancet Diabetes Endocrinol. 3, 191–197 (2015).
171. Gantz, I., Okamoto, T., Ito, Y., Okuyama, K. &
Engel, S. S. Effect of omarigliptin, a novel once-weekly
DPP-4 inhibitor, in Japanese patients with type 2
diabetes: a placebo-and sitagliptin-controlled trial.
Diabetologia 57, S55 (2014).
172. Park, H., Park, C., Kim, Y. & Rascati, K. L. Efficacy and
safety of dipeptidyl peptidase-4 inhibitors in type 2
diabetes: meta-analysis. Ann. Pharmacother. 46,
1453–1469 (2012).
173. Ussher, J. R. & Drucker, D. J. Cardiovascular biology of
the incretin system. Endocr. Rev. 33, 187–215 (2012).
174. Scheen, A. J. Cardiovascular effects of gliptins. Nat.
Rev. Cardiol. 10, 73–84 (2013).
175. Scirica, B. M. et al. Saxagliptin and cardiovascular
outcomes in patients with type 2 diabetes mellitus.
N. Engl. J. Med. 369, 1317–1326 (2013).
176. Udell, J. A. et al. Saxagliptin and cardiovascular outcomes
in patients with type 2 diabetes and moderate or severe
renal impairment: observations from the SAVOR-TIMI 53
trial. Diabetes Care 38, 696–705 (2015).
177. White, W. B. et al. Alogliptin after acute coronary
syndrome in patients with type 2 diabetes. N. Engl.
J. Med. 369, 1327–1335 (2013).
178. Green, J. B. et al. Effect of sitagliptin on cardiovascular
outcomes in type 2 diabetes. N. Engl. J. Med. 373,
232–242 (2015).
179. Li, L. et al. Incretin treatment and risk of pancreatitis
in patients with type 2 diabetes mellitus: systematic
review and meta-analysis of randomised and non-
randomised studies. BMJ 348, g2366 (2014).
180. Roshanov, P. S. & Dennis, B. B. Incretin-based
therapies are associated with acute pancreatitis:
meta-analysis of large randomized controlled trials.
Diabetes Res. Clin. Pract. 110, e13–e17 (2015).
181. Eng, J., Kleinman, W. A., Singh, L., Singh, G. &
Raufman, J. P. Isolation and characterization of
exendin-4, an exendin-3 analogue, from Heloderma
suspectum venom. Further evidence for an exendin
receptor on dispersed acini from guinea pig pancreas.
J. Biol. Chem. 267, 7402–7405 (1992).
182. Barnett, A. Exenatide. Expert. Opin. Pharmacother. 8,
2593–2608 (2007).
183. Drucker, D. J. et al. Exenatide once weekly versus
twice daily for the treatment of type 2 diabetes: a
randomised, open-label, non-inferiority study. Lancet
372, 1240–1250 (2008).
184. Green, B. D. & Flatt, P. R. Incretin hormone mimetics
and analogues in diabetes therapeutics. Best. Pract.
Res. Clin. Endocrinol. Metab. 21, 497–516 (2007).
185. Christensen, M. & Knop, F. K. Once-weekly GLP-1
agonists: how do they differ from exenatide and
liraglutide? Curr. Diab. Rep. 10, 124–132 (2010).
186. Christensen, M., Knop, F. K., Holst, J. J. & Vilsboll, T.
Lixisenatide, a novel GLP-1 receptor agonist for the
treatment of type 2 diabetes mellitus. IDrugs 12,
503–513 (2009).
187. Jimenez-Solem, E., Rasmussen, M. H., Christensen, M.
& Knop, F. K. Dulaglutide, a long-acting GLP-1 analog
fused with an Fc antibody fragment for the potential
treatment of type 2 diabetes. Curr. Opin. Mol. Ther.
12, 790–797 (2010).
188. Nauck, M. A., Vardarli, I., Deacon, C. F., Holst, J. J. &
Meier, J. J. Secretion of glucagon-like peptide-1
(GLP-1) in type 2 diabetes: what is up, what is down?
Diabetologia 54, 10–18 (2011).
189. Kolterman, O. G. et al. Pharmacokinetics,
pharmacodynamics, and safety of exenatide in
patients with type 2 diabetes mellitus. Am. J. Health
Syst. Pharm. 62, 173–181 (2005).
190. Copley, K. et al. Investigation of exenatide elimination
and its in vivo and in vitro degradation. Curr. Drug
Metab. 7, 367–374 (2006).
191. Edwards, C. M. et al. Exendin-4 reduces fasting and
postprandial glucose and decreases energy intake in
healthy volunteers. Am. J. Physiol. Endocrinol. Metab.
281, E155–E161 (2001).
192. Simonsen, L., Holst, J. J. & Deacon, C. F. Exendin-4,
but not glucagon-like peptide-1, is cleared exclusively
by glomerular filtration in anaesthetised pigs.
Diabetologia 49, 706–712 (2006).
193. Linnebjerg, H. et al. Effect of renal impairment on the
pharmacokinetics of exenatide. Br. J. Clin. Pharmacol.
64, 317–327 (2007).
194. Jose, B., Tahrani, A. A., Piya, M. K. & Barnett, A. H.
Exenatide once weekly: clinical outcomes and patient
satisfaction. Patient Prefer. Adherence 4, 313–324
(2010).
195. Elbrønd, B. et al. Pharmacokinetics,
pharmacodynamics, safety, and tolerability of a single-
dose of NN2211, a long-acting glucagon-like peptide 1
derivative, in healthy male subjects. Diabetes Care 25,
1398–1404 (2002).
196. Juhl, C. B. et al. Bedtime administration of NN2211, a
long-acting GLP-1 derivative, substantially reduces
fasting and postprandial glycemia in type 2 diabetes.
Diabetes 51, 424–429 (2002).
197. Agersø, H., Jensen, L. B., Elbrønd, B., Rolan, P. &
Zdravkovic, M. The pharmacokinetics,
pharmacodynamics, safety and tolerability of
NN2211, a new long-acting GLP-1 derivative, in
healthy men. Diabetologia 45, 195–202 (2002).
198. Barnett, A. H. Lixisenatide: evidence for its potential
use in the treatment of type 2 diabetes. Core Evid. 6,
67–79 (2011).
199. Matthews, J. E. et al. Pharmacodynamics,
pharmacokinetics, safety, and tolerability of
albiglutide, a long-acting glucagon-like peptide-1
mimetic, in patients with type 2 diabetes. J. Clin.
Endocrinol. Metab. 93, 4810–4817 (2008).
200. Barrington, P. et al. A 5-week study of the
pharmacokinetics and pharmacodynamics of
LY2189265, a novel, long-acting glucagon-like
peptide-1 analogue, in patients with type 2 diabetes.
Diabetes Obes. Metab. 13, 426–433 (2011).
201. Hurren, K. M. & Pinelli, N. R. Drug–drug interactions
with glucagon-like peptide-1 receptor agonists. Ann.
Pharmacother. 46, 710–717 (2012).
202. Buse, J. B. et al. Effects of exenatide (exendin-4) on
glycemic control over 30 weeks in sulfonylurea-treated
patients with type 2 diabetes. Diabetes Care 27,
2628–2635 (2004).
203. DeFronzo, R. A. et al. Effects of exenatide (exendin-4)
on glycemic control and weight over 30 weeks in
metformin-treated patients with type 2 diabetes.
Diabetes Care 28, 1092–1100 (2005).
204. Kendall, D. M. et al. Effects of exenatide (exendin-4)
on glycemic control over 30 weeks in patients with
type 2 diabetes treated with metformin and a
sulfonylurea. Diabetes Care 28, 1083–1091
(2005).
205. Buse, J. B. et al. Liraglutide once a day versus
exenatide twice a day for type 2 diabetes: a 26-week
randomised, parallel-group, multinational, open-label
trial (LEAD-6). Lancet 374, 39–47 (2009).
206. Garber, A. et al. Liraglutide versus glimepiride
monotherapy for type 2 diabetes (LEAD-3 Mono): a
randomised, 52-week, phase III, double-blind, parallel-
treatment trial. Lancet 373, 473–481 (2009).
207. Marre, M. et al. Liraglutide, a once-daily human GLP-1
analogue, added to a sulphonylurea over 26 weeks
produces greater improvements in glycaemic and
weight control compared with adding rosiglitazone or
placebo in subjects with Type 2 diabetes (LEAD-1 SU).
Diabet. Med. 26, 268–278 (2009).
208. Nauck, M. et al. Efficacy and safety comparison of
liraglutide, glimepiride, and placebo, all in
combination with metformin, in type 2 diabetes: the
LEAD (liraglutide effect and action in diabetes)-2
study. Diabetes Care 32, 84–90 (2009).
209. Zinman, B. et al. Efficacy and safety of the human
glucagon-like peptide-1 analog liraglutide in
combination with metformin and thiazolidinedione in
patients with type 2 diabetes (LEAD-4 Met+TZD).
Diabetes Care 32, 1224–1230 (2009).
210. Bolli, G. B. et al. Efficacy and safety of lixisenatide
once daily versus placebo in people with Type-2
diabetes insufficiently controlled on metformin
(GetGoal-F1). Diabet. Med. 31, 176–184 (2014).
211. Ahrén, B., Leguizamo Dimas, A., Miossec, P.,
Saubadu, S. & Aronson, R. Efficacy and safety of
lixisenatide once-daily morning or evening injections
in type 2 diabetes inadequately controlled on
metformin (GetGoal-M). Diabetes Care 36,
2543–2550 (2013).
212. Pinget, M. et al. Efficacy and safety of lixisenatide
once daily versus placebo in type 2 diabetes
insufficiently controlled on pioglitazone (GetGoal-P).
Diabetes Obes. Metab. 15, 1000–1007 (2013).
213. Riddle, M. C. et al. Adding once-daily lixisenatide for
type 2 diabetes inadequately controlled with newly
initiated and continuously titrated basal insulin
glargine: a 24-week, randomized, placebo-controlled
study (GetGoal-Duo 1). Diabetes Care 36,
2497–2503 (2013).
214. Seino, Y., Min, K. W., Niemoeller, E. & Takami, A.
Randomized, double-blind, placebo-controlled trial of
the once-daily GLP-1 receptor agonist lixisenatide in
Asian patients with type 2 diabetes insufficiently
controlled on basal insulin with or without a
sulfonylurea (GetGoal-L-Asia). Diabetes Obes. Metab.
14, 910–917 (2012).
215. Fonseca, V. A. et al. Efficacy and safety of the once-
daily GLP-1 receptor agonist lixisenatide in
monotherapy: a randomized, double-blind, placebo-
controlled trial in patients with type 2 diabetes
(GetGoal-Mono). Diabetes Care 35, 1225–1231
(2012).
216. Rosenstock, J. et al. Efficacy and safety of lixisenatide
once daily versus exenatide twice daily in type 2
diabetes inadequately controlled on metformin: a
24-week, randomized, open-label, active-controlled
study (GetGoal-X). Diabetes Care 36, 2945–2951
(2013).
217. Ratner, R. E., Rosenstock, J. & Boka, G. Dose-
dependent effects of the once-daily GLP-1 receptor
agonist lixisenatide in patients with type 2 diabetes
inadequately controlled with metformin: a
randomized, double-blind, placebo-controlled trial.
Diabet. Med. 27, 1024–1032 (2010).
218. Bergenstal, R. M. et al. Efficacy and safety of
exenatide once weekly versus sitagliptin or
pioglitazone as an adjunct to metformin for treatment
of type 2 diabetes (DURATION-2): a randomised trial.
Lancet 376, 431–439 (2010).
219. Blevins, T. et al. DURATION-5: exenatide once weekly
resulted in greater improvements in glycemic control
compared with exenatide twice daily in patients with
type 2 diabetes. J. Clin. Endocrinol. Metab. 96,
1301–1310 (2011).
220. Buse, J. B. et al. DURATION-1: exenatide once weekly
produces sustained glycemic control and weight loss
over 52 weeks. Diabetes Care 33, 1255–1261 (2010).
221. Buse, J. B. et al. Exenatide once weekly versus
liraglutide once daily in patients with type 2 diabetes
(DURATION-6): a randomised, open-label study.
Lancet 381, 117–124 (2013).
222. Diamant, M. et al. Exenatide once weekly versus
insulin glargine for type 2 diabetes (DURATION-3):
3-year results of an open-label randomised trial.
Lancet Diabetes Endocrinol. 2, 464–473 (2014).
223. Home, P. D. et al. Efficacy and tolerability of
albiglutide versus placebo or pioglitazone over 1 year
in people with type 2 diabetes currently taking
metformin and glimepiride: HARMONY 5. Diabetes
Obes. Metab. 17, 179–187 (2015).
224. Weissman, P. N. et al. HARMONY 4: randomised
clinical trial comparing once-weekly albiglutide and
insulin glargine in patients with type 2 diabetes
inadequately controlled with metformin with or without
sulfonylurea. Diabetologia 57, 2475–2484 (2014).
225. Reusch, J. et al. Efficacy and safety of once-weekly
glucagon-like peptide 1 receptor agonist albiglutide
(HARMONY 1 trial): 52-week primary endpoint results
from a randomized, double-blind, placebo-controlled
trial in patients with type 2 diabetes mellitus not
controlled on pioglitazone, with or without metformin.
Diabetes Obes. Metab. 16, 1257–1264 (2014).
226. Ahrén, B. et al. HARMONY 3: 104-week randomized,
double-blind, placebo- and active-controlled trial
assessing the efficacy and safety of albiglutide
compared with placebo, sitagliptin, and glimepiride in
patients with type 2 diabetes taking metformin.
Diabetes Care 37, 2141–2148 (2014).
227. Pratley, R. E. et al. Once-weekly albiglutide versus
once-daily liraglutide in patients with type 2 diabetes
inadequately controlled on oral drugs (HARMONY 7):
a randomised, open-label, multicentre, non-inferiority
phase 3 study. Lancet Diabetes Endocrinol. 2,
289–297 (2014).
228. Reaney, M., Yu, M., Lakshmanan, M., Pechtner, V. &
van Brunt, K. Treatment satisfaction in people with
type 2 diabetes mellitus treated with once-weekly
dulaglutide: data from the AWARD-1 and AWARD-3
clinical trials. Diabetes Obes. Metab. 17, 896–903
(2015).
229. Weinstock, R. S. et al. Safety and efficacy of once-
weekly dulaglutide versus sitagliptin after 2 years in
metformin-treated patients with type 2 diabetes
(AWARD-5): a randomized, phase III study. Diabetes
Obes. Metab. 17, 849–858 (2015).
230. Dungan, K. M. et al. Once-weekly dulaglutide versus
once-daily liraglutide in metformin-treated patients
with type 2 diabetes (AWARD-6): a randomised, open-
label, phase 3, non-inferiority trial. Lancet 384,
1349–1357 (2014).
231. Wysham, C. et al. Efficacy and safety of dulaglutide
added onto pioglitazone and metformin versus
exenatide in type 2 diabetes in a randomized
controlled trial (AWARD-1). Diabetes Care 37,
2159–2167 (2014).
232. Umpierrez, G., Tofé Povedano, S., Pérez Manghi, F.,
Shurzinske, L. & Pechtner, V. Efficacy and safety of
dulaglutide monotherapy versus metformin in type 2
diabetes in a randomized controlled trial (AWARD-3).
Diabetes Care 37, 2168–2176 (2014).
233. Skrivanek, Z. et al. Dose-finding results in an adaptive,
seamless, randomized trial of once-weekly dulaglutide
combined with metformin in type 2 diabetes patients
(AWARD-5). Diabetes Obes. Metab. 16, 748–756
(2014).
234. Nauck, M. et al. Efficacy and safety of dulaglutide
versus sitagliptin after 52 weeks in type 2 diabetes in
a randomized controlled trial (AWARD-5). Diabetes
Care 37, 2149–2158 (2014).
235. Russell-Jones, D. et al. Liraglutide versus insulin
glargine and placebo in combination with metformin
and sulfonylurea therapy in type 2 diabetes mellitus
(LEAD-5 met+SU): a randomised controlled trial.
Diabetologia 52, 2046–2055 (2009).
236. Diamant, M. et al. Once weekly exenatide compared
with insulin glargine titrated to target in patients with
type 2 diabetes (DURATION-3): an open-label
randomised trial. Lancet 375, 2234–2243 (2010).
237. Giorgino, F., Benroubi, M., Sun, J. H.,
Zimmermann, A. G. & Pechtner, V. Efficacy and safety
of once-weekly dulaglutide versus insulin glargine in
patients with type 2 diabetes on metformin and
glimepiride (AWARD-2). Diabetes Care 38,
2241–2249 (2015).
238. Meier, J. J. GLP-1 receptor agonists for individualized
treatment of type 2 diabetes mellitus. Nat. Rev.
Endocrinol. 8, 728–742 (2012).
239. Harris, K. B. & McCarty, D. J. Efficacy and tolerability
of glucagon-like peptide-1 receptor agonists in
patients with type 2 diabetes mellitus. Ther. Adv.
Endocrinol. Metab. 6, 3–18 (2015).
240. Klonoff, D. C. et al. Exenatide effects on diabetes,
obesity, cardiovascular risk factors and hepatic
biomarkers in patients with type 2 diabetes treated
for at least 3 years. Curr. Med. Res. Opin. 24,
275–286 (2008).
241. Aroda, V. R. et al. Efficacy of GLP-1 receptor
agonists and DPP-4 inhibitors: meta-analysis and
systematic review. Clin. Ther. 34, 1247–1258
(2012).
242. Buse, J. B. et al. Use of twice-daily exenatide in Basal
insulin-treated patients with type 2 diabetes: a
randomized, controlled trial. Ann. Intern. Med. 154,
103–112 (2011).
243. Rosenstock, J. et al. Baseline factors associated with
glycemic control and weight loss when exenatide twice
daily is added to optimized insulin glargine in patients
with type 2 diabetes. Diabetes Care 35, 955–958
(2012).
244. Deacon, C. F., Mannucci, E. & Ahrén, B. Glycaemic
efficacy of glucagon-like peptide-1 receptor agonists
and dipeptidyl peptidase-4 inhibitors as add-on
therapy to metformin in subjects with type 2 diabetes
— a review and meta analysis. Diabetes Obes. Metab.
14, 762–767 (2012).
245. Henry, R. R. et al. DURATION-1 extension: efficacy
and tolerability of exenatide once weekly (QW) over
6 years in patients with T2DM. Diabetes 63, A247
(2014).
246. Klein, E. et al. DURATION-1 extension: efficacy and
tolerability of exenatide once weekly over 6 years in
patients with type 2 diabetes mellitus. Diabetologia
57, S39 (2014).
247. Garber, A. et al. Liraglutide, a once-daily human
glucagon-like peptide 1 analogue, provides sustained
improvements in glycaemic control and weight for
2 years as monotherapy compared with glimepiride in
patients with type 2 diabetes. Diabetes Obes. Metab.
13, 348–356 (2011).
248. Ahmann, A. et al. Efficacy and safety of liraglutide
versus placebo added to basal insulin analogues (with
or without metformin) in patients with type 2 diabetes:
a randomized, placebo-controlled trial. Diabetes Obes.
Metab. 17, 1056–1064 (2015).
249. Henry, R. R. et al. Efficacy of anthiyperglycemic
therapies and the influence of baseline hemoglobin
A(1C): a meta-analysis of the liraglutide development
program. Endocr. Pract. 17, 906–913 (2011).
250. Tella, S. H. & Rendell, M. S. Glucagon-like polypeptide
agonists in type 2 diabetes mellitus: efficacy and
tolerability, a balance. Ther. Adv. Endocrinol. Metab. 6,
109–134 (2015).
251. Yu Pan, C. et al. Lixisenatide treatment improves
glycaemic control in Asian patients with type 2
diabetes mellitus inadequately controlled on
metformin with or without sulfonylurea: a randomized,
double-blind, placebo-controlled, 24-week trial
(GetGoal-M-Asia). Diabetes Metab. Res. Rev. 30,
726–735 (2014).
252. Rosenstock, J. et al. Beneficial effects of once-daily
lixisenatide on overall and postprandial glycemic levels
without significant excess of hypoglycemia in Type 2
diabetes inadequately controlled on a sulfonylurea
with or without metformin (GetGoal-S). J. Diabetes
Complications 28, 386–392 (2014).
253. Riddle, M. C. et al. Adding once-daily lixisenatide for
type 2 diabetes inadequately controlled by established
basal insulin: a 24-week, randomized, placebo-
controlled comparison (GetGoal-L). Diabetes Care 36,
2489–2496 (2013).
254. Raccah, D., Gourdy, P., Sagnard, L. & Ceriello, A.
Lixisenatide as add-on to oral anti-diabetic therapy: an
effective treatment for glycaemic control with body
weight benefits in type 2 diabetes. Diabetes Metab.
Res. Rev. 30, 742–748 (2014).
255. Schmidt, L. J., Habacher, W., Augustin, T., Krahulec, E.
& Semlitsch, T. A systematic review and meta-analysis
of the efficacy of lixisenatide in the treatment of
patients with type 2 diabetes. Diabetes Obes. Metab.
16, 769–779 (2014).
256. Ahrén, B. et al. Pronounced reduction of postprandial
glucagon by lixisenatide: a meta-analysis of
randomized clinical trials. Diabetes Obes. Metab. 16,
861–868 (2014).
257. Charbonnel, B., Bertolini, M., Tinahones, F. J.,
Domingo, M. P. & Davies, M. Lixisenatide plus basal
insulin in patients with type 2 diabetes mellitus: a
meta-analysis. J. Diabetes Complications 28,
880–886 (2014).
258. Russell-Jones, D. et al. Efficacy and safety of exenatide
once weekly versus metformin, pioglitazone, and
sitagliptin used as monotherapy in drug-naive patients
with type 2 diabetes (DURATION-4): a 26-week double-
blind study. Diabetes Care 35, 252–258 (2012).
259. Davies, M. et al. Once-weekly exenatide versus once-
or twice-daily insulin detemir: randomized, open-label,
clinical trial of efficacy and safety in patients with
type 2 diabetes treated with metformin alone or in
combination with sulfonylureas. Diabetes Care 36,
1368–1376 (2013).
260. Wysham, C. H. et al. Five-year efficacy and safety data
of exenatide once weekly: long-term results from the
DURATION-1 randomized clinical trial. Mayo Clin.
Proc. 90, 356–365 (2015).
261. Rosenstock, J. et al. Potential of albiglutide, a long-
acting GLP-1 receptor agonist, in type 2 diabetes: a
randomized controlled trial exploring weekly, biweekly,
and monthly dosing. Diabetes Care 32, 1880–1886
(2009).
262. Karagiannis, T. et al. Efficacy and safety of once-weekly
glucagon-like peptide 1 receptor agonists for the
management of type 2 diabetes: a systematic review
and meta-analysis of randomized controlled trials.
Diabetes Obes. Metab. 17, 1065–1074 (2015).
263. Rosenstock, J. et al. Advancing basal insulin
replacement in type 2 diabetes inadequately
controlled with insulin glargine plus oral agents: a
comparison of adding albiglutide, a weekly GLP-1
receptor agonist, versus thrice-daily prandial insulin
lispro. Diabetes Care 37, 2317–2325 (2014).
264. Kim, Y. G., Hahn, S., Oh, T. J., Park, K. S. & Cho, Y. M.
Differences in the HbA1c-lowering efficacy of
glucagon-like peptide-1 analogues between Asians
and non-Asians: a systematic review and meta-
analysis. Diabetes Obes. Metab. 16, 900–909
(2014).
265. Vilsbøll, T., Christensen, M., Junker, A. E., Knop, F. K.
& Gluud, L. L. Effects of glucagon-like peptide-1
receptor agonists on weight loss: systematic review
and meta-analyses of randomised controlled trials.
BMJ 344, d7771 (2012).
266. Eng, C., Kramer, C. K., Zinman, B. &
Retnakaran, R. Glucagon-like peptide-1 receptor
agonist and basal insulin combination treatment
for the management of type 2 diabetes: a
systematic review and meta-analysis. Lancet 384,
2228–2234 (2014).
267. Potts, J. E. et al. The effect of glucagon-like peptide 1
receptor agonists on weight loss in type 2 diabetes: a
systematic review and mixed treatment comparison
meta-analysis. PLoS ONE 10, e0126769 (2015).
268. Sun, F. et al. Effect of GLP-1 receptor agonists on waist
circumference among type 2 diabetes patients: a
systematic review and network meta-analysis.
Endocrine 48, 794–803 (2015).
269. Katout, M. et al. Effect of GLP-1 mimetics on blood
pressure and relationship to weight loss and glycemia
lowering: results of a systematic meta-analysis and
meta-regression. Am. J. Hypertens. 27, 130–139
(2014).
270. Sun, F. et al. Impact of GLP-1 receptor agonists on
blood pressure, heart rate and hypertension among
patients with type 2 diabetes: a systematic review and
network meta-analysis. Diabetes Res. Clin. Pract. 110,
26–37 (2015).
271. Ferdinand, K. C. et al. Effects of the once-weekly
glucagon-like peptide-1 receptor agonist dulaglutide
on ambulatory blood pressure and heart rate in
patients with type 2 diabetes mellitus. Hypertension
64, 731–737 (2014).
272. Sun, F. et al. Effect of glucagon-like peptide-1 receptor
agonists on lipid profiles among type 2 diabetes: a
systematic review and network meta-analysis. Clin.
Ther. 37, 225–241 (2015).
273. Egan, A. G. et al. Pancreatic safety of incretin-based
drugs — FDA and EMA assessment. N. Engl. J. Med.
370, 794–797 (2014).
274. Wang, T. et al. Using real-world data to evaluate the
association of incretin-based therapies with risk of
acute pancreatitis: a meta-analysis of 1,324,515
patients from observational studies. Diabetes Obes.
Metab. 17, 32–41 (2015).
275. Giorda, C. B. et al. Incretin-based therapies and acute
pancreatitis risk: a systematic review and meta-
analysis of observational studies. Endocrine 48,
461–471 (2015).
276. Pfeffer, M. A. et al. Lixisenatide in patients with type 2
diabetes and acute coronary syndrome. N. Engl.
J. Med. 373, 2247–2257 (2015).
277. Lønborg, J. et al. Exenatide reduces reperfusion injury
in patients with ST-segment elevation myocardial
infarction. Eur. Heart J. 33, 1491–1499 (2012).
278. Gill, A. et al. Effect of exenatide on heart rate and
blood pressure in subjects with type 2 diabetes
mellitus: a double-blind, placebo-controlled,
randomized pilot study. Cardiovasc. Diabetol. 9, 6
(2010).
279. Trujillo, J. M., Nuffer, W. & Ellis, S. L. GLP-1 receptor
agonists: a review of head-to-head clinical studies.
Ther. Adv. Endocrinol. Metab. 6, 19–28 (2015).
280. Owens, D. R., Monnier, L. & Bolli, G. B. Differential
effects of GLP-1 receptor agonists on components of
dysglycaemia in individuals with type 2 diabetes
mellitus. Diabetes Metab. 39, 485–496 (2013).
281. Meier, J. J. et al. Contrasting effects of lixisenatide
and liraglutide on postprandial glycemic control,
gastric emptying, and safety parameters in patients
with type 2 diabetes on optimized insulin glargine with
or without metformin: a randomized, open-label trial.
Diabetes Care 38, 1263–1273 (2015).
282. Wang, Y. et al. Glucagon-like peptide-1 receptor
agonists versus insulin in inadequately controlled
patients with type 2 diabetes mellitus: a meta-analysis
of clinical trials. Diabetes Obes. Metab. 13, 972–981
(2011).
283. Blonde, L. et al. Once-weekly dulaglutide versus
bedtime insulin glargine, both in combination with
prandial insulin lispro, in patients with type 2
diabetes (AWARD-4): a randomised, open-label,
phase 3, non-inferiority study. Lancet 385,
2057–2066 (2015).
284. Gough, S. C. et al. Efficacy and safety of a fixed-ratio
combination of insulin degludec and liraglutide
(IDegLira) compared with its components given alone:
results of a phase 3, open-label, randomised,
26-week, treat-to-target trial in insulin-naive patients
with type 2 diabetes. Lancet Diabetes Endocrinol. 2,
885–893 (2014).
285. Gough, S. C. et al. One-year efficacy and safety of a
fixed combination of insulin degludec and liraglutide in
patients with type 2 diabetes: results of a 26-week
extension to a 26-week main trial. Diabetes Obes.
Metab. 17, 965–973 (2015).
286. Buse, J. B. et al. Contribution of liraglutide in the
fixed-ratio combination of insulin degludec and
liraglutide (IDegLira). Diabetes Care 37, 2926–2933
(2014).
287. Rosenstock, J. et al. Benefits of a fixed-ratio
formulation of once-daily insulin glargine/lixisenatide
(LixiLan) versus glargine in type 2 diabetes
inadequately controlled on metformin. Diabetologia
57, S108 (2014).
288. [no authors listed.] Sanofi reports positive top-line
results in second pivotal LixiLan phase III study.
SANOFI http://en.sanofi.com/NasdaQ_OMX/local/
press_releases/sanofi_reports_positive_
toplin_1951405_14-09-2015!07_00_00.aspx
(2015).
289. Tahrani, A. A., Barnett, A. H. & Bailey, C. J. SGLT
inhibitors in management of diabetes. Lancet Diabetes
Endocrinol. 1, 140–151 (2013).
290. Scheen, A. J. Pharmacodynamics, efficacy and safety
of sodium-glucose co-transporter type 2 (SGLT2)
inhibitors for the treatment of type 2 diabetes
mellitus. Drugs 75, 33–59 (2015).
291. Wright, E. M., Hirayama, B. A. & Loo, D. F. Active
sugar transport in health and disease. J. Intern. Med.
261, 32–43 (2007).
292. Thorens, B. & Mueckler, M. Glucose transporters in
the 21st Century. Am. J. Physiol. Endocrinol. Metab.
298, E141–E145 (2010).
293. Bailey, C. J. & Day, C. SGLT2 inhibitors: glucuretic
treatment for type 2 diabetes. Br. J. Diabetes Vasc.
Dis. 10, 193–199 (2010).
294. Bailey, C. J. Renal glucose reabsorption inhibitors to
treat diabetes. Trends Pharmacol. Sci. 32, 63–71
(2011).
295. Wright, E. M., Loo, D. D. & Hirayama, B. A. Biology of
human sodium glucose transporters. Physiol. Rev. 91,
733–794 (2011).
296. Wright, E. M. Renal Na+-glucose cotransporters. Am.
J. Physiol. Renal Physiol. 280, F10–F18 (2001).
297. Nauck, M. A. et al. Dapagliflozin versus glipizide as
add-on therapy in patients with type 2 diabetes who
have inadequate glycemic control with metformin: a
randomized, 52-week, double-blind, active-controlled
noninferiority trial. Diabetes Care 34, 2015–2022
(2011).
298. Ferrannini, G. et al. Energy balance after sodium-
glucose cotransporter 2 inhibition. Diabetes Care 38,
1730–1735 (2015).
299. Fulcher, G. et al. Efficacy and safety of canagliflozin
when used in conjunction with incretin-mimetic
therapy in patients with type 2 diabetes. Diabetes
Obes. Metab. 18, 82–91 (2016).
300. Obermeier, M. et al. In vitro characterization and
pharmacokinetics of dapagliflozin (BMS-512148), a
potent sodium-glucose cotransporter type II inhibitor,
in animals and humans. Drug Metab. Dispos. 38,
405–414 (2010).
301. Meng, W. et al. Discovery of dapagliflozin: a potent,
selective renal sodium-dependent glucose
cotransporter 2 (SGLT2) inhibitor for the treatment of
type 2 diabetes. J. Med. Chem. 51, 1145–1149
(2008).
302. Devineni, D. et al. Canagliflozin improves glycaemic
control over 28 days in subjects with type 2 diabetes
not optimally controlled on insulin. Diabetes Obes.
Metab. 14, 539–545 (2012).
303. Heise, T. et al. Safety, tolerability, pharmacokinetics
and pharmacodynamics following 4 weeks’ treatment
with empagliflozin once daily in patients with type 2
diabetes. Diabetes Obes. Metab. 15, 613–621
(2013).
304. Grempler, R. et al. Empagliflozin, a novel selective
sodium glucose cotransporter-2 (SGLT-2) inhibitor:
characterisation and comparison with other SGLT-2
inhibitors. Diabetes Obes. Metab. 14, 83–90
(2012).
305. Jahagirdar, V. & Barnett, A. H. Empagliflozin for the
treatment of type 2 diabetes. Expert Opin.
Pharmacother. 15, 2429–2441 (2014).
306. Scheen, A. J. Drug–drug interactions with sodium-
glucose cotransporters type 2 (SGLT2) inhibitors, new
oral glucose-lowering agents for the management of
type 2 diabetes mellitus. Clin. Pharmacokinet. 53,
295–304 (2014).
307. Ferrannini, E., Ramos, S. J., Salsali, A., Tang, W. &
List, J. F. Dapagliflozin monotherapy in type 2 diabetic
patients with inadequate glycemic control by diet and
exercise: a randomized, double-blind, placebo-
controlled, phase 3 trial. Diabetes Care 33,
2217–2224 (2010).
308. Sun, Y. N., Zhou, Y., Chen, X., Che, W. S. &
Leung, S. W. The efficacy of dapagliflozin combined
with hypoglycaemic drugs in treating type 2 diabetes
mellitus: meta-analysis of randomised controlled trials.
BMJ Open 4, e004619 (2014).
309. Nauck, M. A. et al. Durability of glycaemic efficacy
over 2 years with dapagliflozin versus glipizide as
add-on therapies in patients whose type 2 diabetes
mellitus is inadequately controlled with metformin.
Diabetes Obes. Metab. 16, 1111–1120 (2014).
310. Grandy, S., Hashemi, M., Langkilde, A. M., Parikh, S.
& Sjöström, C. D. Changes in weight loss-related
quality of life among type 2 diabetes mellitus patients
treated with dapagliflozin. Diabetes Obes. Metab. 16,
645–650 (2014).
311. Merovci, A. et al. Dapagliflozin improves muscle
insulin sensitivity but enhances endogenous glucose
production. J. Clin. Invest. 124, 509–514 (2014).
312. Bonner, C. et al. Inhibition of the glucose transporter
SGLT2 with dapagliflozin in pancreatic α cells triggers
glucagon secretion. Nat. Med. 21, 512–517 (2015).
313. Yang, X. P., Lai, D., Zhong, X. Y., Shen, H. P. &
Huang, Y. L. Efficacy and safety of canagliflozin in
subjects with type 2 diabetes: systematic review and
meta-analysis. Eur. J. Clin. Pharmacol. 70,
1149–1158 (2014).
314. Neal, B. et al. Efficacy and safety of canagliflozin, an
inhibitor of sodium-glucose cotransporter 2, when used
in conjunction with insulin therapy in patients with
type 2 diabetes. Diabetes Care 38, 403–411 (2015).
315. Stein, P. et al. Canagliflozin, a sodium glucose
co-transporter 2 inhibitor, reduces post-meal glucose
excursion in patients with type 2 diabetes by a non-
renal mechanism: results of a randomized trial.
Metabolism 63, 1296–1303 (2014).
316. Leiter, L. A. et al. Canagliflozin provides durable
glycemic improvements and body weight reduction
over 104 weeks versus glimepiride in patients with
type 2 diabetes on metformin: a randomized, double-
blind, phase 3 study. Diabetes Care 38, 355–364
(2015).
317. Roden, M. et al. Empagliflozin monotherapy with
sitagliptin as an active comparator in patients with
type 2 diabetes: a randomised, double-blind, placebo-
controlled, phase 3 trial. Lancet Diabetes Endocrinol.
1, 208–219 (2013).
318. Häring, H. U. et al. Empagliflozin as add-on to
metformin in patients with type 2 diabetes: a
24-week, randomized, double-blind, placebo-
controlled trial. Diabetes Care 37, 1650–1659
(2014).
319. Häring, H. U. et al. Empagliflozin as add-on to
metformin plus sulfonylurea in patients with type 2
diabetes: a 24-week randomized, double-blind,
placebo-controlled trial. Diabetes Care 36,
3396–3404 (2013).
320. Kovacs, C. S. et al. Empagliflozin improves glycaemic
and weight control as add-on therapy to pioglitazone
or pioglitazone plus metformin in patients with
type 2 diabetes: a 24-week, randomized, placebo-
controlled trial. Diabetes Obes. Metab. 16,
147–158 (2014).
321. Roden, M. et al. Empagliflozin (EMPA) monotherapy
for ≥76 weeks in drug-naïve patients with type 2
diabetes (T2DM) [abstract 264-OR]. Diabetes 63
(Suppl. 1), A69 (2014).
322. Kovacs, C. S. et al. Empagliflozin as add-on therapy to
pioglitazone with or without metformin in patients
with type 2 diabetes mellitus. Clin. Ther. 37,
1773–1788 (2015).
323. Merker, L. et al. Empagliflozin as add-on to metformin
in people with type 2 diabetes. Diabet. Med. 32,
1555–1567 (2015).
324. Haering, H. U. et al. Empagliflozin as add-on to
metformin plus sulphonylurea in patients with type 2
diabetes. Diabetes Res. Clin. Pract. 110, 82–90
(2015).
VCDNG^5
DQZ^5
VCDNG^5
VCDNG^
325. Ridderstråle, M. et al. Comparison of empagliflozin
VCDNG^5
DQZ^5
DQZ
and glimepiride as add-on to metformin in patients
with type 2 diabetes: a 104-week randomised, active-
controlled, double-blind, phase 3 trial. Lancet
Diabetes Endocrinol. 2, 691–700 (2014).
326. Rosenstock, J. et al. Impact of empagliflozin added on
to basal insulin in type 2 diabetes inadequately
controlled on basal insulin: a 78-week randomized,
double-blind, placebo-controlled trial. Diabetes Obes.
Metab. 17, 936–948 (2015).
327. Rosenstock, J. et al. Improved glucose control with
weight loss, lower insulin doses, and no increased
hypoglycemia with empagliflozin added to titrated
multiple daily injections of insulin in obese
inadequately controlled type 2 diabetes. Diabetes
Care 37, 1815–1823 (2014).
328. Tikkanen, I. et al. Empagliflozin reduces blood
pressure in patients with type 2 diabetes and
hypertension. Diabetes Care 38, 420–428 (2015).
329. Barnett, A. H. et al. Efficacy and safety of
empagliflozin added to existing antidiabetes
treatment in patients with type 2 diabetes and chronic
kidney disease: a randomised, double-blind, placebo-
controlled trial. Lancet Diabetes Endocrinol. 2,
369–384 (2014).
330. Nyirjesy, P. et al. Genital mycotic infections with
canagliflozin, a sodium glucose co-transporter 2
inhibitor, in patients with type 2 diabetes mellitus: a
pooled analysis of clinical studies. Curr. Med. Res.
Opin. 30, 1109–1119 (2014).
331. Hach, T. et al. Empagliflozin improves glycemic
parameters and cardiovascular risk factors in patients
with type 2 diabetes (T2DM): pooled data from four
pivotal Phase III trials [abstract 69-LB]. Diabetes 62
(Suppl. 1A), LB19 (2013).
332. Ptaszynska, A. et al. Safety profile of dapagliflozin for
type 2 diabetes: pooled analysis of clinical studies for
overall safety and rare events. Drug Saf. 37, 815–829
(2014).
333. Sha, S. et al. Effect of the sodium glucose
co-transporter 2 inhibitor canagliflozin on plasma
volume in patients with type 2 diabetes mellitus.
Diabetes Obes. Metab. 16, 1087–1095 (2014).
334. Toto, R. et al. No overall increase in volume depletion
events with empagliflozin (EMPA) in a pooled analysis
of more than 11,000 patients with type 2 diabetes
(T2DM) [abstract SA-PO373]. J. Am. Soc. Nephrol.
24, 711A (2013).
335. Ljunggren, Ö. et al. Dapagliflozin has no effect on
markers of bone formation and resorption or bone
mineral density in patients with inadequately
controlled type 2 diabetes mellitus on metformin.
Diabetes Obes. Metab. 14, 990–999 (2012).
336. Bilezikian, J. P. et al. Evaluation of bone mineral
density and bone biomarkers in patients with type 2
diabetes treated with canagliflozin. J. Clin. Endocrinol.
Metab. 101, 44–51 (2016).
337. Watts, N. B. et al. Effects of canagliflozin on fracture
risk in patients with type 2 diabetes mellitus. J. Clin.
Endocrinol. Metab. 101, 157–166 (2015).
338. Kalra, S., Sahay, R. & Gupta, Y. Sodium glucose
transporter 2 (SGLT2) inhibition and ketogenesis.
Indian J. Endocrinol. Metab. 19, 524–528 (2015).
339. Storgaard, H., Bagger, J. I., Knop, F. K., Vilsbøll, T. &
Rungby, J. Diabetic ketoacidosis in a patient with
type 2 diabetes after initiation of sodium-glucose
cotransporter 2 inhibitor treatment. Basic Clin.
Pharmacol. Toxicol. 118, 168–170 (2016).
340. Peters, A. L. et al. Euglycemic diabetic ketoacidosis: a
potential complication of treatment with sodium-
glucose cotransporter 2 inhibition. Diabetes Care 38,
1687–1693 (2015).
341. Erondu, N., Desai, M., Ways, K. & Meininger, G.
Diabetic ketoacidosis and related events in the
canagliflozin type 2 diabetes clinical program.
Diabetes Care 38, 1680–1686 (2015).
342. Rosenstock, J. & Ferrannini, E. Euglycemic diabetic
ketoacidosis: a predictable, detectable, and
preventable safety concern with SGLT2 inhibitors.
Diabetes Care 38, 1638–1642 (2015).
343. Zinman, B. et al. Empagliflozin, cardiovascular
outcomes, and mortality in type 2 diabetes. N. Engl.
J. Med. 373, 2117–2128 (2015).
344. Holt, R. I., Barnett, A. H. & Bailey, C. J. Bromocriptine:
old drug, new formulation and new indication.
Diabetes Obes. Metab. 12, 1048–1057 (2010).
345. Cincotta, A. H., Meier, A. H. & Cincotta, J. M.
Bromocriptine improves glycaemic control and serum
lipid profile in obese Type 2 diabetic subjects: a new
approach in the treatment of diabetes. Expert. Opin.
Investig. Drugs 8, 1683–1707 (1999).
346. Barnett, A. H., Chapman, C., Gailer, K. & Hayter, C. J.
Effect of bromocriptine on maturity onset diabetes.
Postgrad. Med. J. 56, 11–14 (1980).
347. Lam, C. K., Chari, M. & Lam, T. K. CNS regulation of
glucose homeostasis. Physiology (Bethesda) 24,
159–170 (2009).
348. Liang, W. et al. Efficacy and safety of
bromocriptine-QR in type 2 diabetes: a systematic
review and meta-analysis. Horm. Metab. Res. 47,
805–812 (2015).
349. Gaziano, J. M. et al. Randomized clinical trial of
quick-release bromocriptine among patients with
type 2 diabetes on overall safety and
cardiovascular outcomes. Diabetes Care 33,
1503–1508 (2010).
350. Fonseca, V. A., Handelsman, Y. & Staels, B.
Colesevelam lowers glucose and lipid levels in type 2
diabetes: the clinical evidence. Diabetes Obes. Metab.
12, 384–392 (2010).
351. Rosenstock, J. et al. Efficacy and safety of colesevelam
in combination with pioglitazone in patients with
type 2 diabetes mellitus. Horm. Metab. Res. 46,
943–949 (2014).
352. Younk, L. M., Mikeladze, M. & Davis, S. N. Pramlintide
and the treatment of diabetes: a review of the data
since its introduction. Expert Opin. Pharmacother. 12,
1439–1451 (2011).
353. Henry, R. R. et al. Randomized trial of continuous
subcutaneous delivery of exenatide by ITCA 650
versus twice-daily exenatide injections in metformin-
treated type 2 diabetes. Diabetes Care 36,
2559–2565 (2013).
354. Nauck, M. A. et al. A phase 2, randomized, dose-
finding study of the novel once-weekly human GLP-1
analog, semaglutide, compared with placebo and
open-label liraglutide in patients with type 2 diabetes.
Diabetes Care 39, 231–241 (2016).
355. Biftu, T. et al. Omarigliptin (MK-3102): a novel long-
acting DPP-4 inhibitor for once-weekly treatment of
type 2 diabetes. J. Med. Chem. 57, 3205–3212
(2014).
356. Amin, N. B. et al. Dose-ranging efficacy and safety
study of ertugliflozin, a sodium-glucose co-transporter
2 inhibitor, in patients with type 2 diabetes on a
background of metformin. Diabetes Obes. Metab. 17,
591–598 (2015).
357. Amin, N. B. et al. Blood pressure lowering effect of the
sodium glucose co-transporter (SGLT2) inhibitor,
ertugliflozin, assessed via ambulatory blood pressure
monitoring (ABPM), in patients with type 2 diabetes
and hypertension. Diabetes Obes. Metab. 17,
805–808 (2015).
358. van Gaal, L. & Scheen, A. Weight management in
type 2 diabetes: current and emerging approaches to
treatment. Diabetes Care 38, 1161–1172 (2015).
359. Bailey, C. J. Interpreting adverse signals in diabetes
drug development programs. Diabetes Care 36,
2098–2106 (2013).
Author contributions
All authors researched data for the article, made substantial
contributions to discussions of content and contributed to writ-
ing the article, as well as reviewing and editing the manuscript
before submission.
Competing interests statement
A.A.T. is a clinician scientist supported by the National Institute
for Health Research (NIHR) in the UK. The views expressed in
this publication are those of the author(s) and not necessarily
those of the National Health Service, the NIHR, or the
Department of Health. A.A.T. has received honoraria for lec-
tures and advisory work or support for attending conferences
from Boehringer-Ingelheim, Bristol-Myers Squibb, Eli Lilly,
Novo Nordisk and Sanofi-Aventis. A.A.T. received investiga-
tor-led grant support from the Novo Nordisk Research
Foundation. A.H.B. has received honoraria for lectures and
advisory work from AstraZeneca, Boehringer-Ingelheim, Eli
Lilly, Janssen, MSD, Novartis, Novo Nordisk, Sanofi-Aventis
and Takeda. C.J.B. has undertaken ad-hoc consultancy for
AstraZeneca, Bristol-Myers Squibb, Elcelyx, Eli Lilly, Janssen,
Lexicon, MSD, Novo Nordisk, Roche, Sanofi-Aventis and
Takeda. C.J.B. has delivered continuing medical education pro-
grammes sponsored by AstraZeneca, Boehringer Ingelheim,
Bristol-Myers Squibb, Eli Lilly and MSD, and received travel
or accommodation reimbursement from AstraZeneca and
Bristol-Myers Squibb.
SUPPLEMENTARY INFORMATION
See online article: 5
5